Influence of HPV type on prognosis in patients diagnosed with invasive cervical cancer K. Cuschieri 1 , D.H. Brewster 2 , C. Graham 3 , S. Nicoll 4 , A.R.W. Williams 5 , G.I. Murray 6 , D. Millan 7 , I. Johannessen 8 , A. Hardie 8 and H.A. Cubie 9 1 Scottish HPV Reference Laboratory, Royal Infirmary of Edinburgh, Edinburgh, United Kingdom 2 Scottish Cancer Registry, Information Services Division, NHS National Services Scotland, Edinburgh, United Kingdom 3 Wellcome Trust Clinical Research Facility, University of Edinburgh Western General Hospital, Edinburgh, United Kingdom 4 Department of Pathology, Ninewells Hospital, Dundee, United Kingdom 5 Department of Pathology, Royal Infirmary of Edinburgh, Edinburgh, United Kingdom 6 Department of Pathology, Aberdeen Royal Infirmary, Aberdeen, United Kingdom 7 Department of Pathology, Southern General Hospital, Glasgow, United Kingdom 8 Specialist Virology Centre, Royal Infirmary of Edinburgh, Edinburgh, United Kingdom 9 HPV Research Group, University of Edinburgh, Edinburgh, United Kingdom While much is known about the influence of HPV type on the progression of pre-invasive cervical lesions, the impact of HPV type on cervical cancer prognosis is less evidenced. Thus, we assessed the impact of HPV type on the survival of women diag- nosed with cervical cancer. A total of 370 cases of cervical cancer were assessed. Univariate analysis is presented using Kaplan–Meier survival curves and log-rank statistics and multivariable Cox proportional hazard models were generated using age group, socio-economic deprivation, FIGO stage, differentiation and HPV type. HPV grouping was considered in a number of ways with particular reference to the presence or absence of HPV 16 and/or 18. In the univariate analysis, FIGO, age at diag- nosis and treatment were associated with poorer survival (p < 0.0001) as was absence of HPV 16 and/or 18 (p 5 0.0460). The 25% mortality time in the non-HPV 16/18 vs. HPV16/18 positive group was 615 days and 1,307 days respectively. An unad- justed Cox PH model based HPV16/18 vs. no HPV 16/18 resulted in a hazard ratio of 0.669 (0.450, 0.995). Adjusting for deprivation, FIGO and age group resulted in a hazard ratio of 0.609 (0.395, 0.941) p 5 0.025. These data indicate that cancers associated with HPV 16 and/or 18 do not confer worse survival compared to cancers associated with other types, and may indicate improved survival. Consequently, although HPV vaccine is likely to reduce the incidence of cervical cancer it may not indirectly improve cervical cancer survival by reducing the burden of those cancers caused by HPV16/18. It is now well established that the majority of invasive cervi- cal cancers (ICCs) are associated with HPV 16 and or 18; in a global meta-analysis of nearly 31,000 cases these two types were responsible for between 70 and 76% of ICC in nearly all countries assessed. 1 There is also significant evidence to suggest that HPV 16 and 18 confer a worse prognosis in the context of cervical infection and pre-invasive disease. In the influential study of Khan et al. (2005), the authors found the 10 year cumulative incidence of high-grade cervical lesions (CIN31) in around 20,000 women with normal or low grade cervical smears was 17.2% in HPV 16 positive women, 13.6% in HPV 18 positive women and only 3.0% in women positive for other high risk (HR) HPV types. 2 The observation that HPV 16 is the highest-risk of the high risk types has been shown to be consistent in colpo- scopy populations—and consequently, clinical management algorithms that incorporate HPV genotyping for risk stratifi- cation are being developed, as are high-throughput HPV con- sensus assays that include separate channels for HPV 16 and 18. 3,4 However, although there is now much evidence on the impact of HPV type in the trajectory of pre-invasive disease, there is little information on how it affects prognosis in ICC. Furthermore, existing studies that address the impact of type on survival are heterogeneous, specifically with respect to assay choice (serological vs. molecular), the method of case selection (i.e., whether the cervix was the primary tumor) and whether cervical cancer was confirmed as the primary cause of death (or not). 5–11 In addition, many studies are derived from small sample sets (<50) of patients undergoing specific treatments. Gaining a more comprehensive insight into how HPV type affects survival is important, particularly as we expect that the incidence of HPV 16 and 18 associated ICC will Key words: HPV, genotype, cervical, cancer, survival Grant sponsor: Chief Scientist Office of the Scottish Government DOI: 10.1002/ijc.28902 History: Received 3 Sep 2013; Accepted 3 Apr 2014; Online 17 Apr 2014 Correspondence to: K. Cuschieri, Scottish HPV Reference Laboratory, Royal Infirmary of Edinburgh, Edinburgh, United Kingdom, Tel.: 1[44-131-242-6039], Fax: 1[44-131-242-6008], E-mail: kate.cuschieri@luht.scot.nhs.uk Short Report Int. J. Cancer: 135, 2721–2726 (2014) V C 2014 UICC International Journal of Cancer IJC