Imaging plays a crucial role in identifying breast cancers at a very early stage by means of screening, characterising small tumours in the breast, guiding biopsy of the breast lesion and the axillary lymph nodes thus contributing to staging of the disease thereby providing for choosing an appropriate treatment modality.

Early breast cancers are often detected during screening of breasts which otherwise are not demonstrating lumps or any other signs of tumour on clinical examination. Consequently, these are small tumours with no spread to other structures although not all small tumors can be said to be early cancers as a small percentage of tiny primary breast cancers present clinically or radiologically with metastases.

Stage 0 (ductal carcinoma in situ - DCIS, lobular carcinoma in situ - LCIS and Paget disease of the breast with no tumour) and stage 1 (tumours of size between 1 mm and 20 mm) tumours of the breast with no evidence of spread are considered early breast cancers. Upto 11% of predetermined DCIS on imaging may have an invasive component at the time of biopsy and about a fourth of all DCISs diagnosed on core biopsy may demonstrate invasive ductal carcinoma following surgical excision.

Different imaging modalities help identifying breast cancers early before clinical presentation or before metastases

Mammography, ultrasound and MRI are often used in various combinations to diagnose early breast cancers.

1. Mammography uses X rays; full field digital mammography (FFDM) is replacing the earlier used analog machines in most places. Further improvement is seen in FFDM in the form of digital breast tomosynthesis (3D-DBT), which helps in acquiring images of the entire volume of breast tissue at different angles to overcome the masking effect limitation of 2D images. Contrast enhanced spectral mammography (CESM or angiomammography) with the use of intravenous iodinated contrast administration can increase the sensitivity of mammography in detecting breast cancers, which is almost similar to MRI. Iodinated contrast can also be used during DBT to enhance tumour detection/characterisation.
2. Breast ultrasound with a high resolution transducer (10 to 15 MHz) is often used to detect and localise DCIS, generally in conjunction with mammography. Recent literature has promising results with the use of sono-elastography for characterising the breast lesions by assessing the consistency (hardness/softness) of the lesions. Ultrasound helps in image guided interventions like biopsy or wire/radiotracer localisation of the often non-palpable tumour.
3. MRI is performed for further characterisation of a suspicious lesion detected on mammography and/or ultrasound, particularly when a decision to perform breast conservation surgery is contemplated. MRI is also performed in select high-risk patients as a screening modality. A dedicated breast coil is required; precontrast T1W and post-contrast T1W (dynamic contrast enhancement - DCE), images are most commonly acquired with fat saturation technique. T2W and diffusion weighted images are obtained as well before administration of the gadolinium contrast. These sequences help in determining morphology and cellularity of the lesion respectively. Recent literature has shown some hope regarding the use of spectroscopy images as well as MRI elastography in breast MRI. DCE images can be used to assess the enhancement pattern of the breast lesion with respect to time by plotting time intensity curve on a dedicated MR workstation.

Stage 0 breast cancer

Mammographically, DCIS can be seen as casting type of microcalcifications (50 to 70%) or a soft tissue density (20 to 25%) with or without microcalcifications though rarely it can be seen as an asymmetric density (8%); important point to remember is that not all DCISs calcify and hence mammographic distribution of calcifications underestimates the extent of DCIS. 3D tomosynthesis has a better rate of detection of these lesions by eliminating the masking effect limitation of the 2D FFDM.

A hypoechoic mass with microlobulations and ductal extension with no significant distal acoustic shadowing is the most common feature of a sonographically detected DCIS. It should be understood that not all mammographically detected DCISs can be identified by ultrasound and hence ultrasound is almost always used in conjunction with mammography during the initial evaluation.

DCIS on MRI is generally seen as a non-mass like area of enhancement, most commonly with a segmental or linear distribution and clumped or heterogeneous internal enhancement pattern; clustered ring internal enhancement pattern is most specific for malignancy, usually DCIS.

Stage 1 breast cancer

These are small tumours between 0.1 and 2.0 cm in size, which are often seen as spiculated focal densities on mammography/3D tomosynthesis with architectural distortion with or without calcifications; ultrasound also demonstrates a similar heterogenous/hypoechoic lesion with spiculated margins and distal acoustic shadowing; elastography demonstrates a hard lesion with the ratio of elastography image to grey scale image being more than 1, suggesting tumour infiltration beyond the morphologically visible margins. MRI typically demonstrates a small spiculated enhancing mass demonstrating restricted diffusion on DWI and an early enhancement with wash off on the time intensity curve.

Role of high resolution ultrasound in early breast cancer

Apart from detecting and characterising the primary breast tumour thereby guiding biopsy/intervention, a high resolution ultrasound is often used for assessing the axillary lymph nodes in early breast cancer patients. A lymph node in the ipsilateral axilla is considered to be suspicious for harbouring malignant cells from the breast cancer if any one or more of the following features are seen:

• cortical thickening of more than 3 mm (focal or diffuse)
• hilar effacement
• non-hilar/trans-capsular cortical blood flow
• compression of the hyperechoic medullary region
• markedly hypoechoic cortex
• more than 10 mm short axis diameter of the node
• height length ratio of more than 0.5

Lymph node size criterion is less reliable than the morphological criteria. A well-performed ultrasound examination of the axillary lymph node has a positive predictive value of more than 80% though the negative predictive value is disappointing (about 60%); accuracy is around 67% with sensitivity and specificity being about 54% and 85% respectively. The efficiency of the ultrasound increases significantly with addition of the ultrasound guided biopsy of the lymph node.

Usual clinical pathway in the management of early breast cancer involves re-operative high resolution ultrasound scan of axillary lymph node followed by biopsy when suspicious node is identified; if biopsy demonstrates metastatic deposits in the lymph node, the surgeon performs breast surgery with axillary lymph node dissection. If ultrasound is negative for suspicious nodes, breast surgery with sentinel lymph node excision biopsy is performed using colouring dye or radionucleide method (ROLL – radioguided occult lymph node localisation).

The author, Dr. Govindarajan MJ is an Interventional Radiologist and the Head of Oncoimaging at a prominent Bengaluru hospital.