- Email: [email protected]
The role of HPV on the risk of second primary neoplasia in patients with oropharyngeal carcinoma
Oral Oncology 64 (2017) 37–43
Contents lists available at ScienceDirect
Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology
The role of HPV on the risk of second primary neoplasia in patients with oropharyngeal carcinoma María Martel a, Laia Alemany b,c, Miren Taberna b,d, Marisa Mena b, Sara Tous b, Silvia Bagué e, Xavier Castellsagué b,c,1, Miquel Quer f, Xavier León f,g,⇑ a
Otorhinolaryngology Department, Hospital Moisès Broggi, Sant Joan Despí, Barcelona, Spain Cancer Epidemiology Research Program, Catalan Institute of Oncology (ICO), IDIBELL, L’Hospitalet de Llobregat, Spain CIBER en Epidemiología y Salud Pública (CIBERESP), Barcelona, Spain d Medical Oncology Department, Catalan Institute of Oncology (ICO), L’Hospitalet de Llobregat, University of Barcelona, Barcelona, Spain e Pathology Department, Hospital de Sant Pau, Barcelona, Spain f Otorhinolaryngology Department, Hospital de Sant Pau, Barcelona, Spain g Centro de Investigación Biomédica en Red de Bioingeniería, Biomateriales y Nanomedicina (CIBER-BBN), Madrid, Spain b c
a r t i c l e
i n f o
Article history: Received 8 August 2016 Received in revised form 15 October 2016 Accepted 24 November 2016
Keywords: Oropharyngeal carcinoma Human papillomavirus (HPV) Second primary neoplasm Tobacco Alcohol Head and neck cancer
a b s t r a c t Objectives: It has been reported that patients with HPV-positive oropharyngeal cancer (OPC) have a lower risk of appearance of second primary neoplasm (SPN) than HPV-negative OPC patients. The aim of our study was to analyze the risk of developing SPN in a large group of patients with OPC according to HPV status in the primary tumor. Materials and methods: We included 412 OPC patients treated at our center from 1991 to 2014 for which the HPV DNA positivity was evaluated by PCR in available tumor specimens. HPV DNA positive samples were further tested for HPV E6⁄I mRNA detection and/or p16INK4a immunohistochemistry. We estimated the incidence of SPN in all cancer sites and in cancer sites related to tobacco and alcohol consumption according to the HPV status in the primary tumor. Results: Fifty-one (12.4%) out of 412 OPCs included in the study were HPV-related. Five-year SPN-free survival for HPV-negative versus HPV-positive OPC patients was 57.0% and 89.0% (P < 0.001), respectively. Corresponding estimates for 10-year SPN-free survival were 35.2% versus 78.5% (P < 0.001). When restricting the analyses to tobacco/alcohol-related SPNs, the corresponding survival rates where 62.0% versus 97.6% (P < 0.001) and 42.2% versus 97.6%, (P < 0.001), for 5-year and 10-year survival rates, respectively. HPV status and previous toxic habits might allow classifying patients regarding the risk of tobacco/alcohol-related SPNs. Conclusion: HPV-related OPC patients have a significant lower risk of SPN development, particularly in those locations related to tobacco use or alcohol consumption. Ó 2016 Published by Elsevier Ltd.
Introduction Patients with head and neck squamous cell carcinoma (HNSCC) have a high risk of developing a second primary neoplasm (SPN), located predominantly in sites epidemiologically related with the classic carcinogenic mechanisms (i.e. tobacco and alcohol consumption) that promoted the appearance of the index tumor such as lung, esophagus and again head and neck [1]. Besides the traditional HNSCC pattern of carcinogenesis associated with tobacco ⇑ Corresponding author at: Hospital de la Santa Creu i Sant Pau, Carrer de Sant Quintí, 89, 08026 Barcelona, Spain. E-mail address: [email protected] (X. León). 1 He passed away on June 12th, 2016. http://dx.doi.org/10.1016/j.oraloncology.2016.11.011 1368-8375/Ó 2016 Published by Elsevier Ltd.
use and alcohol consumption, high-risk human papilloma virus (HPV) has been recognized as an emerging carcinogen, particularly in the oropharynx [2]. These viral mediated tumors are epidemiologically, clinically and biologically different from other HNSCC promoted by the carcinogenic action of tobacco and alcohol [3–5]. Studies evaluating the HPV status in OPC patients with serological markers [6], immunohistochemical p16INK4a staining [7] or in situ hybridization (ISH) [8,9] have reported a decreased incidence in SPN among HPV-positive OPC patients. Moreover, several studies have reported lower rates of SPN among patients with OPC than among patients with other head and neck tumors location, probably related to the HPV etiological origin [10,11]. Furthermore, over the last few years, population studies based on the Surveillance, Epidemiology and End Results (SEER) data
38
M. Martel et al. / Oral Oncology 64 (2017) 37–43
base, a program of the National Cancer Institute that collects cancer incidence and survival data from approximately 28% of the U.S. population [12], have found a decreased incidence in second synchronic [13] and metachronic [14] neoplasm in patients having an index tumor located in the oropharynx. Global increasing trends of HPV-related OPC [15], and the lower risk of SPN among these patients might explain the observed decrease in the development of SPN after an oropharyngeal index tumor. The aim of our study was to analyze the risk and associated factors of developing SPN in OPC patients regarding HPV status. HPV positivity was evaluated by HPV DNA PCR and other HPV oncogenic biomarkers such as E6⁄I mRNA or p16INK4a. Material and methods Patient selection We carried out a study in a single institute cohort of OPC patients from which the clinical data, tumor characteristics and follow-up was collected prospectively [16]. Eligible subjects for the present report were those with a pathologically confirmed primary oropharyngeal invasive squamous cell carcinoma treated at our center between 1991–2014, with available formalin-fixed paraffin-embedded (FFPE) sample of the tumor. During the study period, 835 patients with OPC were treated at our center. Four hundred and twelve were eligible and included in the analysis. Non-eligible subjects were mainly those diagnosed at other centers and referred to our hospital for treatment from whom the FFPE was not available (n = 423). HPV status assessment HPV status of OPC patients diagnosed between 1991–2012, was evaluated as part of a retrospective international study, as previously reported [17]. Briefly, at least five sections were obtained for each paraffin block. First and last sections were used for histopathological evaluation, and the in-between ones for DNA and mRNA HPV testing, and p16INK4a immunohistochemistry (IHC). HPV DNA testing was performed in all cases with SPF-10 PCR/DEIA/LiPA25 system. All HPV-DNA positive samples underwent RNA extraction and E6⁄I mRNA detection by type specific RT-PCR, and p16INK4a expression assessment. HPV status in this retrospective component was based on mRNA positivity as gold standard for HPV oncogenic involvement. To note that HPV E6⁄I mRNA was detected in all cases showing an overexpression of P70% of stained cells. For patients diagnosed between 2013 and 2014, HPV status was prospectively evaluated as part of the clinical routine based on HPV DNA detection by real time PCR amplification following instructions of CLART HPV 2 by GENOMICA (GENOMICA S.A.U, Madrid, Spain). The HPV-DNA positive samples were also evaluated for p16INK4a immunohistochemistry under the manufacturer’s standards (p16 antibody BCAM 16123 on autostainer platform and enVision Flex-linker (DAKO) visualization system). A diffuse pattern of more than 50% of the tumor cells with moderate or high intensity was required to define p16 immunopositive protein expression. HPV status in this prospective component was based on p16INK4a overexpression. Variable considerations The definition of the main outcome variable, SPN, was based on the criteria by Warren and Gates [18]. Accordingly, each SPN should be confirmed histologically, the metastatic origin of a previ-
ous tumor should be ruled out, and there should be no submucous connection between the tumors. Head and neck, lung, bladder and esophagus tumors were considered SPN related to tobacco and/or alcohol consumption. HPV-relatedness in the primary OPC was considered when the sample was HPV DNA positive and either mRNA (retrospective cohort) or p16INK4a (prospective cohort) were positive. TNM 7th edition was used to classify the subsite location of OPC tumors: anterior wall, including base of tongue and vallecula; lateral wall, including tonsil, tonsillar fossa and pillars, and glossotonsillar sulci; superior wall, including the soft palate and uvula; and posterior wall [19]. According to the interaction between tobacco and alcohol consumption, we created a combined variable of toxic consumption with three categories: no consumption; moderate consumption (<20 cigarettes/day and/or <80 gr alcohol/day); and severe consumption (P20 cigarettes/day and/or P80 gr alcohol/day). The median follow-up time of patients included in the study was 5.1 years (standard deviation 4.3 years).
Statistical analyses We carried out a comparative analysis between patients included in the study according to OPC HPV status. Overall SPN and tobacco/alcohol-related SPN frequency were analyzed stratified by age, gender, toxics consumption, local and regional disease extension, tumor differentiation grade, and HPV status. Univariate comparison of patient characteristics between groups was performed using the Chi-square test or Fisher’s exact test for categorical variables, as appropriate. SPN-free survival was estimated according to HPV status. These analyses were performed for overall SPN and for those SPN epidemiologically associated with tobacco and alcohol as previously explained. Actuarial survival was estimated using the KaplanMeier method. The log-rank test was used to compare survival functions. We classified the patients according to the risk of developing SPN in locations related to tobacco and alcohol with a recursive partitioning analysis (classification and regression tree model), taking into account the OPC HPV status and toxic consumption. A multivariate analysis was conducted using Cox’s proportional hazard regression. In these analyses we included the following variables: sex, age, T category, N category, OPC subsite, toxics consumption, and HPV status. Toxics and HPV status were included in the models either separately and combined in a single variable following results of the recursive partitioning analysis. The ethics committees of both Hospital de Sant Pau and the Catalan Institute of Oncology approved the study protocols. The study conforms to the principles outlined in the Declaration of Helsinki.
Results Fifty-seven of the 412 OPC (13.8%) were HPV DNA positive. HPV16 was the genotype most frequently found, in 50 of the 57 HPV DNA-positive samples (87.7%). Other genotypes detected were HPV18 (three cases), and HPV33-35-51-58 (one case each). HPV E6⁄I mRNA was evaluated on 43 HPV DNA positive samples of which 39 (90.7%) were positive (from the retrospective cohort). In 14 HPV DNA positive samples for which mRNA detection was not assessed (from the prospective cohort), p16INK4a IHC was performed and 12 (85.7%) were considered positive. HPV-positive OPC was defined as tumor with HPV DNA positivity and at least another positive biomarker, E6⁄I mRNA (n = 39) or p16INK4a IHC (n = 12). Finally, fifty-one (12.4%) of the OPC cases were considered HPV-positive.
39
M. Martel et al. / Oral Oncology 64 (2017) 37–43 Table 1 Characteristics of the patients according to HPV tumor status. N (%a)
%HPV
b
% HPV+c
Pd
Age
<65 years P65 years
281 (68.2) 131 (31.8)
89.0 84.7
11.0 15.3
0.881
Sex
Male Female
375 (91.0) 37 (9.0)
89.6 67.6
10.4 32.4
0.001
Toxics consumption
Never Moderate Severe
26 (6.3) 52 (12.1) 336 (81.6)
38.5 70.0 94.0
61.5 30.0 6.0
<0.001
T category
T1-2 T3-4
199 (48.3) 213 (51.7)
83.9 91.1
16.1 8.9
0.027
N category
N0 N+
191(46.4) 221 (53.6)
93.7 82.4
6.3 17.6
<0.001
OPC subsite
Lateral wall Anterior wall Superior wall Posterior wall
196 (47.6) 132 (32.0) 57 (13.8) 27 (6.6)
83.2 87.9 96.5 100.0
16.8 12.1 3.5 0.0
0.005
Study period
1991–1998 1999–2006 2007–2014
153 (37.2) 125 (30.3) 134 (32.5)
92.2 88.8 81.3
7.8 11.2 18.7
0.019
OPC: Oropharyngeal cancer. a Column %, proportion among all cases. b Row %, proportion of HPV negativity among cases for each variable category. c Row %, proportion of HPV positivity among cases for each variable category. d p-Value from the HPV /HPV+ % comparison.
Table 1 summarizes the clinical characteristics of the patients according to HPV status. HPV-positive patients were more likely to be female, to have less toxics habits, with a lower local tumor extension but a high regional category at diagnosis, and to have tumors located at the lateral and anterior wall of the oropharynx. There was a statistical significant increase in the frequency of patients with HPV-positive tumors throughout the study period (8% in 1991–1998, 11% in 1999–2006, and 19% in the last period 2007–2014, p = 0.019). One hundred and two patients (24.8%) had a history of a previous tumor before the index OPC, of which 90 were from locations related to tobacco use or alcohol consumption (82 head and neck, 4 lung, 3 bladder and 1 esophagus). One hundred and twenty-four patients (30.1%) had a SPN during the follow-up period after the index OPC. Thirty-two patients (7.8%) had a third primary neoplasm and eight patients (1.9%) a fourth one. Supplementary Table 1 describes the location of the SPN and consecutive neoplasms. One hundred of the 124 SPN neoplasms (80.6%) appeared in locations epidemiologically related to tobacco use or alcohol consumption, including head and neck, lung, esophagus and bladder. During the follow-up period, 115 patients with HPV-negative tumors (31.9%), and 9 patients with HPV-positive tumors (17.6%) had a SPN. Considering only the tobacco/alcohol-related SPNs, 97 of the patients with HPV-negative tumors (26.9%), and 3 patients with HPV-positive tumors (5.9%) had a tobacco/alcohol-related SPN. There were significant differences in the appearance of SPNs and tobacco/alcohol-related SPNs according to the HPV status of the tumor (p = 0.038 and p = 0.001, respectively). Those SPN diagnosed in the HPV-positive patients were located in anatomical sites not related to HPV. Table 2 shows the frequency of any SPN and tobacco/alcoholrelated SPNs according to clinical characteristics and OPC HPV status. Patients with early stage tumors, with higher toxic consumption, soft palate subsite, and HPV-negative status increased the risk of developing SPNs and tobacco/alcohol-related SPNs. Fig. 1 shows the SPN and tobacco/alcohol-related SPN-free survival since OPC index tumor diagnosis according to HPV status. Five and 10-year SPN-free survival for HPV-negative OPC patients
was 57.0% (CI 95%: 49.6–64.4%) and 35.2% (CI 95%: 26.4–44.0%) respectively; and 89.0% (CI 95%: 99.3–78.7%) and 78.5% (CI 95%: 61.9–95.1%) for HPV-positive OPC patients (p < 0.001). When focusing on tobacco/alcohol-related SPNs, 5 and 10-year SPN-free survivals for HPV-negative patients was 62.0% (CI 95%: 54.6– 69.4%) and 42.2% (CI 95%: 32.6–51.8%), respectively, and 97.6% (CI 95%: 92.9–100%) and 97.6% (CI 95%: 92.9–100%) for HPVpositive patients (p < 0.001). According to the results of the Cox regression analysis (Table 3), HPV status was the only variable with an independent prognostic value to predict the risk of development a SPN. HPV-negative OPC patients’ risk of developing a SPN was found to be 3.7 higher than HPV-positive OPC patients (CI 95%: 1.6–8.5, p = 0.002). Similar results were found when the multivariate model was elaborated considering tobacco/alcohol-related SPN as the dependent variable. HPV-negative OPC patients risk of developing a tobacco/ alcohol-related SPN was found to be 10.4 higher than HPVpositive OPC patients (CI 95%: 2.4–45.3, p = 0.002). Moreover, HPV-negative OPC patients had higher frequency of previous tumors than HPV-positive patients (26.9% vs 9.8%, p = 0.008). Regarding previous tumors in locations related to tobacco and alcohol, HPV-negative OPC patients maintained this difference in comparison to HPV-positive patients (24.1% vs 5.9%, p = 0.003) (data not shown). A recursive partitioning analysis classified the risk of appearance of a tobacco/alcohol-related SNP in patients with an OPC according to HPV status and toxics consumption (Fig. 2). Grouping terminal nodes with a similar frequency of SPN we defined three categories of patients with different risk of appearance of tobacco/alcohol-related SNP: ‘‘Low”, HPV-positive and never/moderate toxics consumption; ‘‘Intermediate”, HPV-positive and severe toxics consumption or HPV negative and never/moderate toxics consumption; ‘‘High”, HPV-negative and severe toxics consumption. Fig. 3 shows the tobacco and alcohol related SPN-free survival according to the categories as defined by the recursive partitioning analysis. There were significant differences in the tobacco/alcoholrelated SPN-free survival according the categories (p < 0.001). We repeated the multivariate analysis considering the appearance of tobacco/alcohol-related SNP as the dependent variable
40
M. Martel et al. / Oral Oncology 64 (2017) 37–43
Table 2 Frequency of second primary neoplasm (SPN) and second primary neoplasm in locations related to tobacco and alcohol (SPN⁄) according to several epidemiologic and clinical variables. N
% SPN
P
% SPN⁄
P
Age
<65 years P65 years
281 131
31.7 26.7
0.307
26.3 19.8
0.153
Sex
Male Female
375 37
30.7 24.3
0.422
25.1 16.2
0.231
Toxics consumption
Never Moderate Severe
26 50 336
30.8 14.0 32.4
0.030
11.5 10.0 27.4
0.008
T category
T1-2 T3-4
199 213
35.7 24.9
0.017
28.6 20.2
0.045
N category
N0 N+
191 221
36.1 24.9
0.013
28.8 20.4
0.046
OPC subsite
Lateral wall Anterior wall Superior wall Posterior wall
196 132 57 27
27.6 28.0 47.4 22.2
0.021
21.9 20.5 43.9 28.5
0.003
OPC HPV status
Negative Positive
361 51
31.9 17.3
0.038
26.9 5.9
0.001
OPC: Oropharyngeal cancer.
HPV-positive tumors and moderate or non-history of previous toxic habits as the reference category, patients with HPV-positive tumors and severe toxic consumption or patients with HPVnegative tumors and moderate or non-history of toxic habits had a 4.5 higher risk of appearance of tobacco/alcohol-related SNP, and patients with HPV-negative tumors and a severe history of toxics use had a 13.2 higher risk of appearance of tobacco/ alcohol-related SNP. Discussion
Fig. 1. Any SPN-free survival (A) and tobacco/alcohol-related SPN-free survival (B) according to HPV-related status in the primary OPC tumor (numbers in parenthesis in the number at risk section refers to number of events-deaths for each period of time).
and including in the model a combined variable of toxics and HPV status taking into account the categories obtained with the recursive partitioning analysis (Table 4). Considering the patients with
According to our results, HPV-related OPC patients have lower frequency of previous tumors before index OPC and a significant decreased incidence of SPN than HPV-negative OPC patients. These differences were even more significant if we focused only on previous tumors and SPN epidemiologically related to tobacco use or alcohol consumption as lung, esophagus, bladder and head and neck tumors. Our result confirm the concept proposed by Diaz et al. that oncogenic alterations in HPV-related tumors is spatially limited to the location where they arise [20]. Previous studies have already described a declined trend in incidence for SPN appearance among HPV-positive OPC patients. In a US multicenter study conducted in 323 locally advanced OPC patients with a 63% rate of HPV-positivity determined by ISH, Ang et al. [9] described that the cumulative incidence of SPN was significantly lower among patients with HPV-positive tumors, largely because of lower rates of smoking-related cancers. Saito et al. [8] carried out a study on 150 Japanese OPC patients and reported a 31.3% prevalence of HPV-positive tumors evaluated by DNA-ISH; patients with HPV-positive OPC had a significant lower risk of developing SPN (OR 0.288, CI 95%: 0.102–0.810). Furthermore, on a study conducted in 182 OPC patients at the MD Anderson Cancer Center (Texas, USA), where HPV was determined via serologic methods (43.9% HPV-seropositive patients), Peck et al. [6] reported that HPV-seropositive patients had a higher second primary tumor-free survival than HPV-seronegative patients (p = 0.051). Finally, Xu et al. [7] performed in Alberta (Canada) a study including 199 OPC patients. They used p16INK4a IHC to evaluate the HPV status, finding a rate of 59.3% HPV-positive OPC; the incidence of upper aerodigestive tract SPN (including lung, esophagus and head and neck) for p16INK4a positive patients was significantly lower than for p16INK4a negative patients (P < 0.001). However, there
41
M. Martel et al. / Oral Oncology 64 (2017) 37–43
Table 3 Multivariate analyses considering the time until the appearance of a second primary neoplasm (SPN) or a second primary neoplasm in locations related to tobacco and alcohol (SPN⁄) as the dependent variable. SPN⁄
SPN a
Crude HR (95%CI)
P
HR (95%CI)
P
b
Crude HR (95%CI)
P
HRa (95%CI)
Pb
Age
<65 years P65 years
0.7 (0.5–1.1) 1
0.106
0.7 (0.4–1.0) 1
0.061
0.8 (0.5–1.3) 1
0.394
0.7 (0.4–1.1) 1
0.169
Sex
Male Female
1.6 (0.8–3.2) 1
0.151
1.3 (0.6–2.6) 1
0.521
2.0 (0.9–4.6) 1
0.073
1.4 (0.6–3.3) 1
0.448
Toxics consumption
No Moderate Severe
1 0.8 (0.3–2.1) 1.6 (0.8–3.3)
0.051
1 0.5 (0.2–1.3) 0.8 (0.4–1.8)
0.255
1 1.4 (0.3–6.0) 3.5 (1.1–11.1)
0.004
1 0.7 (0.2–2.9) 1.2 (0.4–4.2)
0.378
T category
T1-2 T3-4
1 1.1 (0.8–1.6)
0.527
1 1.1 (0.8–1.7)
0.536
1 1.1 (0.7–1.7)
0.613
1 1.2 (0.7–1.8)
0.543
N category
N0 N+
1 0.7 (0.5–1.1)
0.104
1 0.98 (0.6–1.5)
0.913
1 0.8 (0.5–1.1)
0.188
1 1.1 (0.7–1.8)
0.658
OPC subsite
Lateral wall Anterior wall Superior wall Posterior wall
1 1.0 (0.7–1.6) 1.3 (0.8–2.2) 0.9 (0.3–2.4)
0.684
1 0.9 (0.6–1.4) 1.1 (0.7–1.9) 0.6 (0.2–1.8)
0.702
1 0.9 (0.6–1.5) 1.6 (0.9–2.6) 0.8 (0.3–2.6)
0.288
1 0.8 (0.5–1.3) 1.4 (0.8–2.5) 0.6 (0.2–1.9)
0.254
HPV status
Positive Negative
1 3.7 (1.8–7.6)
<0.001
1 3.7 (1.6–8.3)
<0.001
1 11.9 (2.9–48.5)
<0.001
1 10.4 (2.4–45.2)
<0.001
OPC: Oropharyngeal cancer. a HR adjusted by Sex, Age, Toxics consumption, T category, N category, OPC subsite and HPV status. b Log-likelihood ratio test p-value.
Fig. 2. Results of the recursive partitioning analysis. Classification of the risk of appearance of a tobacco/alcohol-related second primary neoplasm (SNP⁄) in patients with an OPC according to HPV status and toxics consumption.
were no differences in non-upper aerodigestive tract SPN according the p16INK4a status. A major concern among all the studies described above is that HPV detection technique with ISH, serology or p16INK4a IHC have sensitivity and specificity limitations. One important variable to be considered when analyzing the risk of appearance of SPN is previous tobacco use and alcohol consumption. Peck et al. [6] reported that HPV-seropositive ever smokers patients had a SPN free survival more similar to that of HPV-seronegative patients to that of HPV-seropositive never smokers. Xu et al. [7] described that HPV-positive non-smoker patients have a lower risk of SPN development than HPV-positive smokers. Finally, Gan et al. [10] found that patients with an OPC with a typical HPV phenotype, including non-smokers, had a very
low incidence of SPN, while OPC patients with a classic squamous cell carcinoma phenotype, including those patients aged P60 years, current smokers, and smokers with P20 pack-years of exposure, had a SPN rate similar to that of patients with index non-OPC. Unlike previous studies where tobacco use was the only toxic habit described, for our analysis we have also considered alcohol consumption. Mounting evidence indicates an association between tobacco use and alcohol consumption [21,22], and that both toxics interact increasing the risk of OPC appearance [23]. Our univariate analysis indicated a significant association between toxics consumption and SPN development. However, this association disappeared when HPV status was included in the multivariate
42
M. Martel et al. / Oral Oncology 64 (2017) 37–43
Fig. 3. Tobacco/alcohol-related SPN free survival according to risk categories obtained with the recursive partitioning analysis (numbers in parenthesis in the number at risk section refers to number of events-deaths for each period of time).
analysis. The loss of toxic habits prognostic value would be explained by the differences between tobacco use and alcohol consumption according to patients’ HPV status (Table 2). A recursive partitioning analysis classified our patients in three groups according to the different risk of appearance of a SPN related to tobacco or alcohol. The group with the lowest risk was composed of HPV-positive patients with moderate or non-history of previous toxic habits. HPV-positive patients with severe tobacco use and/or alcohol consumption have an intermediate SPN rate, similar to HPV-negative patients with moderate or non-history of toxic habits. Finally, the major risk of SPN development was achieved by HPV-negative patients with a severe history of toxics use. These findings might have an impact on OPC patient’s follow-up protocols. One of the objectives of the follow-up of patients with a HNSCC is the early diagnosis of secondary metachronic neoplasm, arising more often in the upper-aerodigestive tract [24]. For this reason, follow-up visit should include an exhaustive physical
examination of head and neck mucosa area and a periodic lung screening by image techniques. If our results could be validated, it might be justified to limit HPV-related OPC patients’ follow-up period. Given the low risk of SPN development on HPV-related OPC patients without a history of toxic habits, systematic followup could be limited when the risk of recurrence of the primary tumor decreases. The strength of the present study is to evaluate a large group of patients treated in a single institution, with their clinical reports prospectively collected, and with a long median follow-up period reached. That might guarantee the quality of our data. The most important limitations in our study are related to tobacco use and alcohol consumption evaluation. In our database tobacco use and alcohol consumption are described based on the history of the toxic habit use as null, moderate or severe. Information regarding the period of consumption, if patient quit before the index OPC diagnosis and data on continued smoking and alcohol consumption after treatment was not available. Other of the limitations is the low rate of HPV-positive OPC patients included in the analysis. HPV-related OPC rates in our country are low maybe due to the still high prevalence of smoking habits and differential sexual behaviour in comparison with other geographic regions with high incidence of HPV-positive OPC like US or North European countries. Beside our results, two additional studies have been performed in Spanish population describing a HPV-positive OPC prevalence from 3.2% [25] to 26.7% [26] but in a lower number of cases. Epidemiologic studies in Spain evidenced a trend in increased OPC incidence, despite concomitant declines in incidence for other head and neck locations [27]. Furthermore, we found an increase in HPV-positive proportion of OPC patients throughout the study period. Conclusion HPV-related OPC patients have a significant lower risk of SPN development, particularly and more strongly in cancer sites related to tobacco and alcohol consumption. HPV status and previous toxic habits allows classifying patients regarding the risk of SPN development. This information might have an impact in the design of follow-up protocols for HPV-related OPC patients.
Table 4 Multivariate analyses considering the time until the appearance of a second primary neoplasm in locations related to tobacco and alcohol (SPN⁄) as the dependent variable, including a combined variable for toxics and HPV status according to the recursive partitioning analysis. SPN⁄ Crude HR (95%CI)
P
HRa (95%CI)
Pb
Age
<65 years P65 years
0.8 (0.5–1.3) 1
0.394
0.7 (0.4–1.1) 1
0.112
Sex
Male Female
2.0 (0.9–4.6) 1
0.073
1.3 (0.5–3.0) 1
0.581
Combined variable of toxics and HPV statusc
Low risk Intermediate risk High risk
1 5.6 (0.7–45.2) 14.5 (2.0–104.3)
<0.001
1 4.5 (0.6–36.7) 13.2 (1.8–96.6)
<0.001
T category
T1-2 T3-4
1 1.1 (0.7–1.7)
0.613
1 1.2 (0.7–1.8)
0.510
N category
N0 N+
1 0.8 (0.5–1.1)
0.188
1 1.0 (0.7–1.7)
0.831
OPC subsite
Lateral wall Anterior wall Superior wall Posterior wall
1 0.9 (0.6–1.5) 1.6 (0.9–2.6) 0.8 (0.3–2.6)
0.288
1 0.8 (0.5–1.4) 1.5 (0.8–2.6) 0.6 (0.2–2.1)
0.303
OPC: Oropharyngeal cancer. a HR adjusted by Sex, Age, Combination of Toxics Consumption and HPV status, T category, N category, OPC subsite. b Log-likelihood ratio test p-value. c Combined variable of toxics and HPV status according to RPA analysis: Low risk, HPV-positive and never/moderate toxics consumption; Intermediate risk, HPV-positive and severe toxics consumption or HPV-negative and never/moderate toxics consumption; High risk, HPV-negative and severe toxics consumption.
M. Martel et al. / Oral Oncology 64 (2017) 37–43
Conflict of interest Authors certify that there is no conflict of interest with any financial organization regarding the material discussed in the manuscript. Acknowledgements This work was supported by grants from the Instituto de Salud Carlos III-ISCIII (Spanish Government) cofunded by FEDER funds/ European Regional Development Fund (ERDF) – a way to build Europe (FIS PI11/02096, PI14/01819 and PI15/00378), from ISCIII-Spanish Society of Medical Oncology (Rio Hortega-SEOM grant to MT), and from the Asociación Española Contra el Cáncer (personal grant to LA). Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.oraloncology. 2016.11.011. References [1] Leon X, Quer M, Diez S, Orus C, Lopez-Pousa A, Burgues J. Second neoplasm in patients with head and neck cancer. Head Neck 1999;21:204–10. [2] Weinberger PM, Yu Z, Haffty BG, Kowalski D, Harigopal M, Brandsma J, et al. Molecular classification identifies a subset of human papillomavirus– associated oropharyngeal cancers with favorable prognosis. J Clin Oncol 2006;24:736–47. [3] D’Souza G, Kreimer AR, Viscidi R, Pawlita M, Fakhry C, Koch WM, et al. Casecontrol study of human papillomavirus and oropharyngeal cancer. N England J Med 2007;356:1944–56. [4] Benson E, Li R, Eisele D, Fakhry C. The clinical impact of HPV tumor status upon head and neck squamous cell carcinomas. Oral Oncol 2014;50:565–74. [5] Rampias T, Sasaki C, Psyrri A. Molecular mechanisms of HPV induced carcinogenesis in head and neck. Oral Oncol 2014;50:356–63. [6] Peck BW, Dahlstrom KR, Gan SJ, Caywood W, Li G, Wei Q, et al. Low risk of second primary malignancies among never smokers with human papillomavirus-associated index oropharyngeal cancers. Head Neck 2013;35:794–9. [7] Xu CC, Biron VL, Puttagunta L, Seikaly H. HPV status and second primary tumours in oropharyngeal squamous cell carcinoma. J Otolaryngol Head Neck Surg 2013;42:36. [8] Saito Y, Ebihara Y, Ushiku T, Omura G, Kobayashi K, Ando M, et al. Negative human papillomavirus status and excessive alcohol consumption are significant risk factors for second primary malignancies in Japanese patients with oropharyngeal carcinoma. Jpn J Clin Oncol 2014;44:564–9. [9] Ang KK, Harris J, Wheeler R, Weber R, Rosenthal DI, Nguyen-Tan PF, et al. Human papillomavirus and survival of patients with oropharyngeal cancer. N England J Med 2010;363:24–35.
43
[10] Gan SJ, Dahlstrom KR, Peck BW, Caywood W, Li G, Wei Q, et al. Incidence and pattern of second primary malignancies in patients with index oropharyngeal cancers versus index nonoropharyngeal head and neck cancers. Cancer 2013;119:2593–601. [11] Jung YS, Lim J, Jung KW, Ryu J, Won YJ. Metachronous second primary malignancies after head and neck cancer in a Korean Cohort (1993–2010). PLoS ONE 2015;10:e0134160. [12] Surveillance E, End Results (SEER) Program Research Data, National Cancer Institute.. [13] Jain KS, Sikora AG, Baxi SS, Morris LG. Synchronous cancers in patients with head and neck cancer: risks in the era of human papillomavirus-associated oropharyngeal cancer. Cancer 2013;119:1832–7. [14] Morris LG, Sikora AG, Patel SG, Hayes RB, Ganly I. Second primary cancers after an index head and neck cancer: subsite-specific trends in the era of human papillomavirus-associated oropharyngeal cancer. J Clin Oncol 2011;29:739–46. [15] Chaturvedi AK, Engels EA, Anderson WF, Gillison ML. Incidence trends for human papillomavirus-related and -unrelated oral squamous cell carcinomas in the United States. J Clin Oncol 2008;26:612–9. [16] Leon X, Orus C, Quer M. Design, maintenance, and exploitation of an oncologic database for patients with malignant tumors of the head and neck. Acta Otorrinolaringol Esp 2002;53:185–90. [17] Castellsague X, Alemany L, Quer M, Halec G, Quiros B, Tous S, et al. HPV involvement in head and neck cancers: comprehensive assessment of biomarkers in 3680 patients. J Natl Cancer Inst 2016;108(6):djv403. [18] Warren S, Gates O. Multiple malignant tumors: a survey of literature and statistical study. Am J Cancer 1932;51:1358–414. [19] UICC International Union Against Cancer. TNM classification of malignant tumours. In: Sobin LH, Wittekind CH, editors. 7th ed., Blackwell Publishing Ltd.; 2009 [20] Diaz DA, Reis IM, Weed DT, Elsayyad N, Samuels M, Abramowitz MC. Head and neck second primary cancer rates in the human papillomavirus era: a population-based analysis. Head Neck 2016;38(Suppl 1):E873–83. [21] John U, Meyer C, Rumpf HJ, Schumann A, Thyrian JR, Hapke U. Strength of the relationship between tobacco smoking, nicotine dependence and the severity of alcohol dependence syndrome criteria in a population-based sample. Alcohol Alcohol 2003;38:606–12. [22] Room R. Smoking and drinking as complementary behaviours. Biomed Pharmacother 2004;58:111–5. [23] Blot WJ, McLaughlin JK, Winn DM, Austin DF, Greenberg RS, Preston-Martin S, et al. Smoking and drinking in relation to oral and pharyngeal cancer. Cancer Res 1988;48:3282–7. [24] Leon X, Quer M, Orus C, del Prado Venegas M. Can cure be achieved in patients with head and neck carcinomas? The problem of second neoplasm. Expert Rev Anticancer Ther 2001;1:125–33. [25] Rodrigo JP, Heideman DA, Garcia-Pedrero JM, Fresno MF, Brakenhoff RH, Diaz Molina JP, et al. Time trends in the prevalence of HPV in oropharyngeal squamous cell carcinomas in northern Spain (1990–2009). Int J Cancer 2014;134:487–92. [26] Cerezo L, Lopez C, de la Torre A, Suarez D, Hervas A, Ruiz A, et al. Incidence of human papillomavirus-related oropharyngeal cancer and outcomes after chemoradiation in a population of heavy smokers. Head Neck 2014;36:782–6. [27] de Souza DL, Perez MM, Curado MP. Predicted incidence of oral cavity, oropharyngeal, laryngeal, and hypopharyngeal cancer in Spain and implications for cancer control. Cancer Epidemiol 2011;35:510–4.
Contents lists available at ScienceDirect
Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology
The role of HPV on the risk of second primary neoplasia in patients with oropharyngeal carcinoma María Martel a, Laia Alemany b,c, Miren Taberna b,d, Marisa Mena b, Sara Tous b, Silvia Bagué e, Xavier Castellsagué b,c,1, Miquel Quer f, Xavier León f,g,⇑ a
Otorhinolaryngology Department, Hospital Moisès Broggi, Sant Joan Despí, Barcelona, Spain Cancer Epidemiology Research Program, Catalan Institute of Oncology (ICO), IDIBELL, L’Hospitalet de Llobregat, Spain CIBER en Epidemiología y Salud Pública (CIBERESP), Barcelona, Spain d Medical Oncology Department, Catalan Institute of Oncology (ICO), L’Hospitalet de Llobregat, University of Barcelona, Barcelona, Spain e Pathology Department, Hospital de Sant Pau, Barcelona, Spain f Otorhinolaryngology Department, Hospital de Sant Pau, Barcelona, Spain g Centro de Investigación Biomédica en Red de Bioingeniería, Biomateriales y Nanomedicina (CIBER-BBN), Madrid, Spain b c
a r t i c l e
i n f o
Article history: Received 8 August 2016 Received in revised form 15 October 2016 Accepted 24 November 2016
Keywords: Oropharyngeal carcinoma Human papillomavirus (HPV) Second primary neoplasm Tobacco Alcohol Head and neck cancer
a b s t r a c t Objectives: It has been reported that patients with HPV-positive oropharyngeal cancer (OPC) have a lower risk of appearance of second primary neoplasm (SPN) than HPV-negative OPC patients. The aim of our study was to analyze the risk of developing SPN in a large group of patients with OPC according to HPV status in the primary tumor. Materials and methods: We included 412 OPC patients treated at our center from 1991 to 2014 for which the HPV DNA positivity was evaluated by PCR in available tumor specimens. HPV DNA positive samples were further tested for HPV E6⁄I mRNA detection and/or p16INK4a immunohistochemistry. We estimated the incidence of SPN in all cancer sites and in cancer sites related to tobacco and alcohol consumption according to the HPV status in the primary tumor. Results: Fifty-one (12.4%) out of 412 OPCs included in the study were HPV-related. Five-year SPN-free survival for HPV-negative versus HPV-positive OPC patients was 57.0% and 89.0% (P < 0.001), respectively. Corresponding estimates for 10-year SPN-free survival were 35.2% versus 78.5% (P < 0.001). When restricting the analyses to tobacco/alcohol-related SPNs, the corresponding survival rates where 62.0% versus 97.6% (P < 0.001) and 42.2% versus 97.6%, (P < 0.001), for 5-year and 10-year survival rates, respectively. HPV status and previous toxic habits might allow classifying patients regarding the risk of tobacco/alcohol-related SPNs. Conclusion: HPV-related OPC patients have a significant lower risk of SPN development, particularly in those locations related to tobacco use or alcohol consumption. Ó 2016 Published by Elsevier Ltd.
Introduction Patients with head and neck squamous cell carcinoma (HNSCC) have a high risk of developing a second primary neoplasm (SPN), located predominantly in sites epidemiologically related with the classic carcinogenic mechanisms (i.e. tobacco and alcohol consumption) that promoted the appearance of the index tumor such as lung, esophagus and again head and neck [1]. Besides the traditional HNSCC pattern of carcinogenesis associated with tobacco ⇑ Corresponding author at: Hospital de la Santa Creu i Sant Pau, Carrer de Sant Quintí, 89, 08026 Barcelona, Spain. E-mail address: [email protected] (X. León). 1 He passed away on June 12th, 2016. http://dx.doi.org/10.1016/j.oraloncology.2016.11.011 1368-8375/Ó 2016 Published by Elsevier Ltd.
use and alcohol consumption, high-risk human papilloma virus (HPV) has been recognized as an emerging carcinogen, particularly in the oropharynx [2]. These viral mediated tumors are epidemiologically, clinically and biologically different from other HNSCC promoted by the carcinogenic action of tobacco and alcohol [3–5]. Studies evaluating the HPV status in OPC patients with serological markers [6], immunohistochemical p16INK4a staining [7] or in situ hybridization (ISH) [8,9] have reported a decreased incidence in SPN among HPV-positive OPC patients. Moreover, several studies have reported lower rates of SPN among patients with OPC than among patients with other head and neck tumors location, probably related to the HPV etiological origin [10,11]. Furthermore, over the last few years, population studies based on the Surveillance, Epidemiology and End Results (SEER) data
38
M. Martel et al. / Oral Oncology 64 (2017) 37–43
base, a program of the National Cancer Institute that collects cancer incidence and survival data from approximately 28% of the U.S. population [12], have found a decreased incidence in second synchronic [13] and metachronic [14] neoplasm in patients having an index tumor located in the oropharynx. Global increasing trends of HPV-related OPC [15], and the lower risk of SPN among these patients might explain the observed decrease in the development of SPN after an oropharyngeal index tumor. The aim of our study was to analyze the risk and associated factors of developing SPN in OPC patients regarding HPV status. HPV positivity was evaluated by HPV DNA PCR and other HPV oncogenic biomarkers such as E6⁄I mRNA or p16INK4a. Material and methods Patient selection We carried out a study in a single institute cohort of OPC patients from which the clinical data, tumor characteristics and follow-up was collected prospectively [16]. Eligible subjects for the present report were those with a pathologically confirmed primary oropharyngeal invasive squamous cell carcinoma treated at our center between 1991–2014, with available formalin-fixed paraffin-embedded (FFPE) sample of the tumor. During the study period, 835 patients with OPC were treated at our center. Four hundred and twelve were eligible and included in the analysis. Non-eligible subjects were mainly those diagnosed at other centers and referred to our hospital for treatment from whom the FFPE was not available (n = 423). HPV status assessment HPV status of OPC patients diagnosed between 1991–2012, was evaluated as part of a retrospective international study, as previously reported [17]. Briefly, at least five sections were obtained for each paraffin block. First and last sections were used for histopathological evaluation, and the in-between ones for DNA and mRNA HPV testing, and p16INK4a immunohistochemistry (IHC). HPV DNA testing was performed in all cases with SPF-10 PCR/DEIA/LiPA25 system. All HPV-DNA positive samples underwent RNA extraction and E6⁄I mRNA detection by type specific RT-PCR, and p16INK4a expression assessment. HPV status in this retrospective component was based on mRNA positivity as gold standard for HPV oncogenic involvement. To note that HPV E6⁄I mRNA was detected in all cases showing an overexpression of P70% of stained cells. For patients diagnosed between 2013 and 2014, HPV status was prospectively evaluated as part of the clinical routine based on HPV DNA detection by real time PCR amplification following instructions of CLART HPV 2 by GENOMICA (GENOMICA S.A.U, Madrid, Spain). The HPV-DNA positive samples were also evaluated for p16INK4a immunohistochemistry under the manufacturer’s standards (p16 antibody BCAM 16123 on autostainer platform and enVision Flex-linker (DAKO) visualization system). A diffuse pattern of more than 50% of the tumor cells with moderate or high intensity was required to define p16 immunopositive protein expression. HPV status in this prospective component was based on p16INK4a overexpression. Variable considerations The definition of the main outcome variable, SPN, was based on the criteria by Warren and Gates [18]. Accordingly, each SPN should be confirmed histologically, the metastatic origin of a previ-
ous tumor should be ruled out, and there should be no submucous connection between the tumors. Head and neck, lung, bladder and esophagus tumors were considered SPN related to tobacco and/or alcohol consumption. HPV-relatedness in the primary OPC was considered when the sample was HPV DNA positive and either mRNA (retrospective cohort) or p16INK4a (prospective cohort) were positive. TNM 7th edition was used to classify the subsite location of OPC tumors: anterior wall, including base of tongue and vallecula; lateral wall, including tonsil, tonsillar fossa and pillars, and glossotonsillar sulci; superior wall, including the soft palate and uvula; and posterior wall [19]. According to the interaction between tobacco and alcohol consumption, we created a combined variable of toxic consumption with three categories: no consumption; moderate consumption (<20 cigarettes/day and/or <80 gr alcohol/day); and severe consumption (P20 cigarettes/day and/or P80 gr alcohol/day). The median follow-up time of patients included in the study was 5.1 years (standard deviation 4.3 years).
Statistical analyses We carried out a comparative analysis between patients included in the study according to OPC HPV status. Overall SPN and tobacco/alcohol-related SPN frequency were analyzed stratified by age, gender, toxics consumption, local and regional disease extension, tumor differentiation grade, and HPV status. Univariate comparison of patient characteristics between groups was performed using the Chi-square test or Fisher’s exact test for categorical variables, as appropriate. SPN-free survival was estimated according to HPV status. These analyses were performed for overall SPN and for those SPN epidemiologically associated with tobacco and alcohol as previously explained. Actuarial survival was estimated using the KaplanMeier method. The log-rank test was used to compare survival functions. We classified the patients according to the risk of developing SPN in locations related to tobacco and alcohol with a recursive partitioning analysis (classification and regression tree model), taking into account the OPC HPV status and toxic consumption. A multivariate analysis was conducted using Cox’s proportional hazard regression. In these analyses we included the following variables: sex, age, T category, N category, OPC subsite, toxics consumption, and HPV status. Toxics and HPV status were included in the models either separately and combined in a single variable following results of the recursive partitioning analysis. The ethics committees of both Hospital de Sant Pau and the Catalan Institute of Oncology approved the study protocols. The study conforms to the principles outlined in the Declaration of Helsinki.
Results Fifty-seven of the 412 OPC (13.8%) were HPV DNA positive. HPV16 was the genotype most frequently found, in 50 of the 57 HPV DNA-positive samples (87.7%). Other genotypes detected were HPV18 (three cases), and HPV33-35-51-58 (one case each). HPV E6⁄I mRNA was evaluated on 43 HPV DNA positive samples of which 39 (90.7%) were positive (from the retrospective cohort). In 14 HPV DNA positive samples for which mRNA detection was not assessed (from the prospective cohort), p16INK4a IHC was performed and 12 (85.7%) were considered positive. HPV-positive OPC was defined as tumor with HPV DNA positivity and at least another positive biomarker, E6⁄I mRNA (n = 39) or p16INK4a IHC (n = 12). Finally, fifty-one (12.4%) of the OPC cases were considered HPV-positive.
39
M. Martel et al. / Oral Oncology 64 (2017) 37–43 Table 1 Characteristics of the patients according to HPV tumor status. N (%a)
%HPV
b
% HPV+c
Pd
Age
<65 years P65 years
281 (68.2) 131 (31.8)
89.0 84.7
11.0 15.3
0.881
Sex
Male Female
375 (91.0) 37 (9.0)
89.6 67.6
10.4 32.4
0.001
Toxics consumption
Never Moderate Severe
26 (6.3) 52 (12.1) 336 (81.6)
38.5 70.0 94.0
61.5 30.0 6.0
<0.001
T category
T1-2 T3-4
199 (48.3) 213 (51.7)
83.9 91.1
16.1 8.9
0.027
N category
N0 N+
191(46.4) 221 (53.6)
93.7 82.4
6.3 17.6
<0.001
OPC subsite
Lateral wall Anterior wall Superior wall Posterior wall
196 (47.6) 132 (32.0) 57 (13.8) 27 (6.6)
83.2 87.9 96.5 100.0
16.8 12.1 3.5 0.0
0.005
Study period
1991–1998 1999–2006 2007–2014
153 (37.2) 125 (30.3) 134 (32.5)
92.2 88.8 81.3
7.8 11.2 18.7
0.019
OPC: Oropharyngeal cancer. a Column %, proportion among all cases. b Row %, proportion of HPV negativity among cases for each variable category. c Row %, proportion of HPV positivity among cases for each variable category. d p-Value from the HPV /HPV+ % comparison.
Table 1 summarizes the clinical characteristics of the patients according to HPV status. HPV-positive patients were more likely to be female, to have less toxics habits, with a lower local tumor extension but a high regional category at diagnosis, and to have tumors located at the lateral and anterior wall of the oropharynx. There was a statistical significant increase in the frequency of patients with HPV-positive tumors throughout the study period (8% in 1991–1998, 11% in 1999–2006, and 19% in the last period 2007–2014, p = 0.019). One hundred and two patients (24.8%) had a history of a previous tumor before the index OPC, of which 90 were from locations related to tobacco use or alcohol consumption (82 head and neck, 4 lung, 3 bladder and 1 esophagus). One hundred and twenty-four patients (30.1%) had a SPN during the follow-up period after the index OPC. Thirty-two patients (7.8%) had a third primary neoplasm and eight patients (1.9%) a fourth one. Supplementary Table 1 describes the location of the SPN and consecutive neoplasms. One hundred of the 124 SPN neoplasms (80.6%) appeared in locations epidemiologically related to tobacco use or alcohol consumption, including head and neck, lung, esophagus and bladder. During the follow-up period, 115 patients with HPV-negative tumors (31.9%), and 9 patients with HPV-positive tumors (17.6%) had a SPN. Considering only the tobacco/alcohol-related SPNs, 97 of the patients with HPV-negative tumors (26.9%), and 3 patients with HPV-positive tumors (5.9%) had a tobacco/alcohol-related SPN. There were significant differences in the appearance of SPNs and tobacco/alcohol-related SPNs according to the HPV status of the tumor (p = 0.038 and p = 0.001, respectively). Those SPN diagnosed in the HPV-positive patients were located in anatomical sites not related to HPV. Table 2 shows the frequency of any SPN and tobacco/alcoholrelated SPNs according to clinical characteristics and OPC HPV status. Patients with early stage tumors, with higher toxic consumption, soft palate subsite, and HPV-negative status increased the risk of developing SPNs and tobacco/alcohol-related SPNs. Fig. 1 shows the SPN and tobacco/alcohol-related SPN-free survival since OPC index tumor diagnosis according to HPV status. Five and 10-year SPN-free survival for HPV-negative OPC patients
was 57.0% (CI 95%: 49.6–64.4%) and 35.2% (CI 95%: 26.4–44.0%) respectively; and 89.0% (CI 95%: 99.3–78.7%) and 78.5% (CI 95%: 61.9–95.1%) for HPV-positive OPC patients (p < 0.001). When focusing on tobacco/alcohol-related SPNs, 5 and 10-year SPN-free survivals for HPV-negative patients was 62.0% (CI 95%: 54.6– 69.4%) and 42.2% (CI 95%: 32.6–51.8%), respectively, and 97.6% (CI 95%: 92.9–100%) and 97.6% (CI 95%: 92.9–100%) for HPVpositive patients (p < 0.001). According to the results of the Cox regression analysis (Table 3), HPV status was the only variable with an independent prognostic value to predict the risk of development a SPN. HPV-negative OPC patients’ risk of developing a SPN was found to be 3.7 higher than HPV-positive OPC patients (CI 95%: 1.6–8.5, p = 0.002). Similar results were found when the multivariate model was elaborated considering tobacco/alcohol-related SPN as the dependent variable. HPV-negative OPC patients risk of developing a tobacco/ alcohol-related SPN was found to be 10.4 higher than HPVpositive OPC patients (CI 95%: 2.4–45.3, p = 0.002). Moreover, HPV-negative OPC patients had higher frequency of previous tumors than HPV-positive patients (26.9% vs 9.8%, p = 0.008). Regarding previous tumors in locations related to tobacco and alcohol, HPV-negative OPC patients maintained this difference in comparison to HPV-positive patients (24.1% vs 5.9%, p = 0.003) (data not shown). A recursive partitioning analysis classified the risk of appearance of a tobacco/alcohol-related SNP in patients with an OPC according to HPV status and toxics consumption (Fig. 2). Grouping terminal nodes with a similar frequency of SPN we defined three categories of patients with different risk of appearance of tobacco/alcohol-related SNP: ‘‘Low”, HPV-positive and never/moderate toxics consumption; ‘‘Intermediate”, HPV-positive and severe toxics consumption or HPV negative and never/moderate toxics consumption; ‘‘High”, HPV-negative and severe toxics consumption. Fig. 3 shows the tobacco and alcohol related SPN-free survival according to the categories as defined by the recursive partitioning analysis. There were significant differences in the tobacco/alcoholrelated SPN-free survival according the categories (p < 0.001). We repeated the multivariate analysis considering the appearance of tobacco/alcohol-related SNP as the dependent variable
40
M. Martel et al. / Oral Oncology 64 (2017) 37–43
Table 2 Frequency of second primary neoplasm (SPN) and second primary neoplasm in locations related to tobacco and alcohol (SPN⁄) according to several epidemiologic and clinical variables. N
% SPN
P
% SPN⁄
P
Age
<65 years P65 years
281 131
31.7 26.7
0.307
26.3 19.8
0.153
Sex
Male Female
375 37
30.7 24.3
0.422
25.1 16.2
0.231
Toxics consumption
Never Moderate Severe
26 50 336
30.8 14.0 32.4
0.030
11.5 10.0 27.4
0.008
T category
T1-2 T3-4
199 213
35.7 24.9
0.017
28.6 20.2
0.045
N category
N0 N+
191 221
36.1 24.9
0.013
28.8 20.4
0.046
OPC subsite
Lateral wall Anterior wall Superior wall Posterior wall
196 132 57 27
27.6 28.0 47.4 22.2
0.021
21.9 20.5 43.9 28.5
0.003
OPC HPV status
Negative Positive
361 51
31.9 17.3
0.038
26.9 5.9
0.001
OPC: Oropharyngeal cancer.
HPV-positive tumors and moderate or non-history of previous toxic habits as the reference category, patients with HPV-positive tumors and severe toxic consumption or patients with HPVnegative tumors and moderate or non-history of toxic habits had a 4.5 higher risk of appearance of tobacco/alcohol-related SNP, and patients with HPV-negative tumors and a severe history of toxics use had a 13.2 higher risk of appearance of tobacco/ alcohol-related SNP. Discussion
Fig. 1. Any SPN-free survival (A) and tobacco/alcohol-related SPN-free survival (B) according to HPV-related status in the primary OPC tumor (numbers in parenthesis in the number at risk section refers to number of events-deaths for each period of time).
and including in the model a combined variable of toxics and HPV status taking into account the categories obtained with the recursive partitioning analysis (Table 4). Considering the patients with
According to our results, HPV-related OPC patients have lower frequency of previous tumors before index OPC and a significant decreased incidence of SPN than HPV-negative OPC patients. These differences were even more significant if we focused only on previous tumors and SPN epidemiologically related to tobacco use or alcohol consumption as lung, esophagus, bladder and head and neck tumors. Our result confirm the concept proposed by Diaz et al. that oncogenic alterations in HPV-related tumors is spatially limited to the location where they arise [20]. Previous studies have already described a declined trend in incidence for SPN appearance among HPV-positive OPC patients. In a US multicenter study conducted in 323 locally advanced OPC patients with a 63% rate of HPV-positivity determined by ISH, Ang et al. [9] described that the cumulative incidence of SPN was significantly lower among patients with HPV-positive tumors, largely because of lower rates of smoking-related cancers. Saito et al. [8] carried out a study on 150 Japanese OPC patients and reported a 31.3% prevalence of HPV-positive tumors evaluated by DNA-ISH; patients with HPV-positive OPC had a significant lower risk of developing SPN (OR 0.288, CI 95%: 0.102–0.810). Furthermore, on a study conducted in 182 OPC patients at the MD Anderson Cancer Center (Texas, USA), where HPV was determined via serologic methods (43.9% HPV-seropositive patients), Peck et al. [6] reported that HPV-seropositive patients had a higher second primary tumor-free survival than HPV-seronegative patients (p = 0.051). Finally, Xu et al. [7] performed in Alberta (Canada) a study including 199 OPC patients. They used p16INK4a IHC to evaluate the HPV status, finding a rate of 59.3% HPV-positive OPC; the incidence of upper aerodigestive tract SPN (including lung, esophagus and head and neck) for p16INK4a positive patients was significantly lower than for p16INK4a negative patients (P < 0.001). However, there
41
M. Martel et al. / Oral Oncology 64 (2017) 37–43
Table 3 Multivariate analyses considering the time until the appearance of a second primary neoplasm (SPN) or a second primary neoplasm in locations related to tobacco and alcohol (SPN⁄) as the dependent variable. SPN⁄
SPN a
Crude HR (95%CI)
P
HR (95%CI)
P
b
Crude HR (95%CI)
P
HRa (95%CI)
Pb
Age
<65 years P65 years
0.7 (0.5–1.1) 1
0.106
0.7 (0.4–1.0) 1
0.061
0.8 (0.5–1.3) 1
0.394
0.7 (0.4–1.1) 1
0.169
Sex
Male Female
1.6 (0.8–3.2) 1
0.151
1.3 (0.6–2.6) 1
0.521
2.0 (0.9–4.6) 1
0.073
1.4 (0.6–3.3) 1
0.448
Toxics consumption
No Moderate Severe
1 0.8 (0.3–2.1) 1.6 (0.8–3.3)
0.051
1 0.5 (0.2–1.3) 0.8 (0.4–1.8)
0.255
1 1.4 (0.3–6.0) 3.5 (1.1–11.1)
0.004
1 0.7 (0.2–2.9) 1.2 (0.4–4.2)
0.378
T category
T1-2 T3-4
1 1.1 (0.8–1.6)
0.527
1 1.1 (0.8–1.7)
0.536
1 1.1 (0.7–1.7)
0.613
1 1.2 (0.7–1.8)
0.543
N category
N0 N+
1 0.7 (0.5–1.1)
0.104
1 0.98 (0.6–1.5)
0.913
1 0.8 (0.5–1.1)
0.188
1 1.1 (0.7–1.8)
0.658
OPC subsite
Lateral wall Anterior wall Superior wall Posterior wall
1 1.0 (0.7–1.6) 1.3 (0.8–2.2) 0.9 (0.3–2.4)
0.684
1 0.9 (0.6–1.4) 1.1 (0.7–1.9) 0.6 (0.2–1.8)
0.702
1 0.9 (0.6–1.5) 1.6 (0.9–2.6) 0.8 (0.3–2.6)
0.288
1 0.8 (0.5–1.3) 1.4 (0.8–2.5) 0.6 (0.2–1.9)
0.254
HPV status
Positive Negative
1 3.7 (1.8–7.6)
<0.001
1 3.7 (1.6–8.3)
<0.001
1 11.9 (2.9–48.5)
<0.001
1 10.4 (2.4–45.2)
<0.001
OPC: Oropharyngeal cancer. a HR adjusted by Sex, Age, Toxics consumption, T category, N category, OPC subsite and HPV status. b Log-likelihood ratio test p-value.
Fig. 2. Results of the recursive partitioning analysis. Classification of the risk of appearance of a tobacco/alcohol-related second primary neoplasm (SNP⁄) in patients with an OPC according to HPV status and toxics consumption.
were no differences in non-upper aerodigestive tract SPN according the p16INK4a status. A major concern among all the studies described above is that HPV detection technique with ISH, serology or p16INK4a IHC have sensitivity and specificity limitations. One important variable to be considered when analyzing the risk of appearance of SPN is previous tobacco use and alcohol consumption. Peck et al. [6] reported that HPV-seropositive ever smokers patients had a SPN free survival more similar to that of HPV-seronegative patients to that of HPV-seropositive never smokers. Xu et al. [7] described that HPV-positive non-smoker patients have a lower risk of SPN development than HPV-positive smokers. Finally, Gan et al. [10] found that patients with an OPC with a typical HPV phenotype, including non-smokers, had a very
low incidence of SPN, while OPC patients with a classic squamous cell carcinoma phenotype, including those patients aged P60 years, current smokers, and smokers with P20 pack-years of exposure, had a SPN rate similar to that of patients with index non-OPC. Unlike previous studies where tobacco use was the only toxic habit described, for our analysis we have also considered alcohol consumption. Mounting evidence indicates an association between tobacco use and alcohol consumption [21,22], and that both toxics interact increasing the risk of OPC appearance [23]. Our univariate analysis indicated a significant association between toxics consumption and SPN development. However, this association disappeared when HPV status was included in the multivariate
42
M. Martel et al. / Oral Oncology 64 (2017) 37–43
Fig. 3. Tobacco/alcohol-related SPN free survival according to risk categories obtained with the recursive partitioning analysis (numbers in parenthesis in the number at risk section refers to number of events-deaths for each period of time).
analysis. The loss of toxic habits prognostic value would be explained by the differences between tobacco use and alcohol consumption according to patients’ HPV status (Table 2). A recursive partitioning analysis classified our patients in three groups according to the different risk of appearance of a SPN related to tobacco or alcohol. The group with the lowest risk was composed of HPV-positive patients with moderate or non-history of previous toxic habits. HPV-positive patients with severe tobacco use and/or alcohol consumption have an intermediate SPN rate, similar to HPV-negative patients with moderate or non-history of toxic habits. Finally, the major risk of SPN development was achieved by HPV-negative patients with a severe history of toxics use. These findings might have an impact on OPC patient’s follow-up protocols. One of the objectives of the follow-up of patients with a HNSCC is the early diagnosis of secondary metachronic neoplasm, arising more often in the upper-aerodigestive tract [24]. For this reason, follow-up visit should include an exhaustive physical
examination of head and neck mucosa area and a periodic lung screening by image techniques. If our results could be validated, it might be justified to limit HPV-related OPC patients’ follow-up period. Given the low risk of SPN development on HPV-related OPC patients without a history of toxic habits, systematic followup could be limited when the risk of recurrence of the primary tumor decreases. The strength of the present study is to evaluate a large group of patients treated in a single institution, with their clinical reports prospectively collected, and with a long median follow-up period reached. That might guarantee the quality of our data. The most important limitations in our study are related to tobacco use and alcohol consumption evaluation. In our database tobacco use and alcohol consumption are described based on the history of the toxic habit use as null, moderate or severe. Information regarding the period of consumption, if patient quit before the index OPC diagnosis and data on continued smoking and alcohol consumption after treatment was not available. Other of the limitations is the low rate of HPV-positive OPC patients included in the analysis. HPV-related OPC rates in our country are low maybe due to the still high prevalence of smoking habits and differential sexual behaviour in comparison with other geographic regions with high incidence of HPV-positive OPC like US or North European countries. Beside our results, two additional studies have been performed in Spanish population describing a HPV-positive OPC prevalence from 3.2% [25] to 26.7% [26] but in a lower number of cases. Epidemiologic studies in Spain evidenced a trend in increased OPC incidence, despite concomitant declines in incidence for other head and neck locations [27]. Furthermore, we found an increase in HPV-positive proportion of OPC patients throughout the study period. Conclusion HPV-related OPC patients have a significant lower risk of SPN development, particularly and more strongly in cancer sites related to tobacco and alcohol consumption. HPV status and previous toxic habits allows classifying patients regarding the risk of SPN development. This information might have an impact in the design of follow-up protocols for HPV-related OPC patients.
Table 4 Multivariate analyses considering the time until the appearance of a second primary neoplasm in locations related to tobacco and alcohol (SPN⁄) as the dependent variable, including a combined variable for toxics and HPV status according to the recursive partitioning analysis. SPN⁄ Crude HR (95%CI)
P
HRa (95%CI)
Pb
Age
<65 years P65 years
0.8 (0.5–1.3) 1
0.394
0.7 (0.4–1.1) 1
0.112
Sex
Male Female
2.0 (0.9–4.6) 1
0.073
1.3 (0.5–3.0) 1
0.581
Combined variable of toxics and HPV statusc
Low risk Intermediate risk High risk
1 5.6 (0.7–45.2) 14.5 (2.0–104.3)
<0.001
1 4.5 (0.6–36.7) 13.2 (1.8–96.6)
<0.001
T category
T1-2 T3-4
1 1.1 (0.7–1.7)
0.613
1 1.2 (0.7–1.8)
0.510
N category
N0 N+
1 0.8 (0.5–1.1)
0.188
1 1.0 (0.7–1.7)
0.831
OPC subsite
Lateral wall Anterior wall Superior wall Posterior wall
1 0.9 (0.6–1.5) 1.6 (0.9–2.6) 0.8 (0.3–2.6)
0.288
1 0.8 (0.5–1.4) 1.5 (0.8–2.6) 0.6 (0.2–2.1)
0.303
OPC: Oropharyngeal cancer. a HR adjusted by Sex, Age, Combination of Toxics Consumption and HPV status, T category, N category, OPC subsite. b Log-likelihood ratio test p-value. c Combined variable of toxics and HPV status according to RPA analysis: Low risk, HPV-positive and never/moderate toxics consumption; Intermediate risk, HPV-positive and severe toxics consumption or HPV-negative and never/moderate toxics consumption; High risk, HPV-negative and severe toxics consumption.
M. Martel et al. / Oral Oncology 64 (2017) 37–43
Conflict of interest Authors certify that there is no conflict of interest with any financial organization regarding the material discussed in the manuscript. Acknowledgements This work was supported by grants from the Instituto de Salud Carlos III-ISCIII (Spanish Government) cofunded by FEDER funds/ European Regional Development Fund (ERDF) – a way to build Europe (FIS PI11/02096, PI14/01819 and PI15/00378), from ISCIII-Spanish Society of Medical Oncology (Rio Hortega-SEOM grant to MT), and from the Asociación Española Contra el Cáncer (personal grant to LA). Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.oraloncology. 2016.11.011. References [1] Leon X, Quer M, Diez S, Orus C, Lopez-Pousa A, Burgues J. Second neoplasm in patients with head and neck cancer. Head Neck 1999;21:204–10. [2] Weinberger PM, Yu Z, Haffty BG, Kowalski D, Harigopal M, Brandsma J, et al. Molecular classification identifies a subset of human papillomavirus– associated oropharyngeal cancers with favorable prognosis. J Clin Oncol 2006;24:736–47. [3] D’Souza G, Kreimer AR, Viscidi R, Pawlita M, Fakhry C, Koch WM, et al. Casecontrol study of human papillomavirus and oropharyngeal cancer. N England J Med 2007;356:1944–56. [4] Benson E, Li R, Eisele D, Fakhry C. The clinical impact of HPV tumor status upon head and neck squamous cell carcinomas. Oral Oncol 2014;50:565–74. [5] Rampias T, Sasaki C, Psyrri A. Molecular mechanisms of HPV induced carcinogenesis in head and neck. Oral Oncol 2014;50:356–63. [6] Peck BW, Dahlstrom KR, Gan SJ, Caywood W, Li G, Wei Q, et al. Low risk of second primary malignancies among never smokers with human papillomavirus-associated index oropharyngeal cancers. Head Neck 2013;35:794–9. [7] Xu CC, Biron VL, Puttagunta L, Seikaly H. HPV status and second primary tumours in oropharyngeal squamous cell carcinoma. J Otolaryngol Head Neck Surg 2013;42:36. [8] Saito Y, Ebihara Y, Ushiku T, Omura G, Kobayashi K, Ando M, et al. Negative human papillomavirus status and excessive alcohol consumption are significant risk factors for second primary malignancies in Japanese patients with oropharyngeal carcinoma. Jpn J Clin Oncol 2014;44:564–9. [9] Ang KK, Harris J, Wheeler R, Weber R, Rosenthal DI, Nguyen-Tan PF, et al. Human papillomavirus and survival of patients with oropharyngeal cancer. N England J Med 2010;363:24–35.
43
[10] Gan SJ, Dahlstrom KR, Peck BW, Caywood W, Li G, Wei Q, et al. Incidence and pattern of second primary malignancies in patients with index oropharyngeal cancers versus index nonoropharyngeal head and neck cancers. Cancer 2013;119:2593–601. [11] Jung YS, Lim J, Jung KW, Ryu J, Won YJ. Metachronous second primary malignancies after head and neck cancer in a Korean Cohort (1993–2010). PLoS ONE 2015;10:e0134160. [12] Surveillance E, End Results (SEER) Program Research Data, National Cancer Institute.