Sunil Lakhani

The recognition that ADH is biologically very similar to DCIS has led to a policy of excising all ADH lesions diagnosed on core biopsies at breast screening assessment

The demonstration that LCIS is a clonal neoplastic proliferation with the same molecular phenotype as invasive lobular carcinoma, and the subsequent dissemination of this knowledge through publication (Lakhani SR 2003, Lancet) and conference presentations has also changed the management of LCIS. It is accepted that it is a non-obligate precursor, rather than a hyperplastic lesion.

The recognition of the pleomorphic variant of lobular is now changing the way this proliferation is managed. Like other lobular lesions, it is multifocal/bilateral but a more aggressive variant. There is considerable ongoing discussion on the international stage regarding management and my work has played a role in this policy change.

We have also shown that columnar cell lesions, which are increasingly being diagnosed in the screening program, are clonal lesions with a progressive relationship to low grade DCIS (Simpson PT et al 2005 Am J Surg Pathol). We have shown that morphological atypia is associated with molecular profiles similar to that of DCIS and hence patients with atypical lesions on core biopsy are now subjected to further excision. It has been shown subsequently that many of these patients have worse pathology on further resection.

My expertise, in diagnostic and molecular biology of pre-invasive breast disease has also been acknowledged in being appointed Head of Breast Screen Pathology, North Brisbane as well as being a member of the Breast Screen Quality (‘Q’) group in Queensland.

My work as part of the BCLC has demonstrated that BRCA1-related breast cancers are high grade, highly proliferative tumours and occurred in young women. Hence there was an immediate impact on breast screening protocols. I was invited to many meetings to discuss the pathology in relation to the following: (i) should mammography be used in young women and how effective would it be in dense breasts? (ii) should another modality such as MRI be considered? (iii) irrespective of the modality, should the screening interval be shortened to reduce the risk that patients would present with interval cancers?

All these issues have been addressed subsequent to the publications – in particular, I was member of the steering group that published the experience of using MRI in young women at high risk of breast cancer (Leach et al 2005 MARIBS The Lancet). MRI is now used in many countries to screen high-risk women. The screening interval has been reduced to 1 year as many patients did present with interval cancers as predicted by the pathology.

In extending the original work on morphology, my group has also demonstrated that a large proportion of BRCA1associated cancers are ER negative and ‘basal-like’. This led us to propose that this phenotype could be used to improve the risk estimation and carrier probability of women seen in family cancer clinics (Lakhani et al 2005 Clin Can Res).

We have now published a number of papers supporting the use of pathology data to improve risk estimation as well as improving the likelihood of unclassified variants (UVs) being pathogenic. This work has been done with previous BCLC collaborators (Prof Doug Easton) as well as with Georgia Chenevix-Trench and Amanda Spurdle at QIMR (Chenevix-Trench et al 2006 Can Res, Lovelock et al 2007 Br Can Res, Spurdle et al 2008 JCO, Spurdle et al 2009 Hum Mut, Mavaddat et al 2010 Br Can Res) This approach is now being incorporated into the genetics clinics.