Literature search results and characteristics of included studies
The search strategy identified a total of 2618 records, of which 855 corresponded to IS and 1763 to NIS. Following removal of duplicate records, records with lack of published articles and/or congress abstracts, and studies not fulfilling the PICOTS criteria, a total of 62 IS and 19 NIS were included in the evidence synthesis (Fig. 1).
[Figure 1 to be inserted here]
All included studies (N=81) provided an HPV prevalence of LA and/or RM HNC captured between 1st of January 2010 and 31st of December 2020 and were used for addressing the primary study objective. Of the included studies, 43 IS and 9 NIS reported data on prevalence of HPV specifically for OPC, and were thus used to address the secondary outcome of interest.
Characteristics of the studies included in the evidence synthesis and outcomes of interest derived from each study are presented in Table 1 and Additional file 3. Of the IS, 42 (67.7%) were single-arm and 20 (32.3%) were multi-arm; of the latter 17 (27.5%) were randomized. Of the NIS, 13 (68.4%) were retrospective, two (10.5%) were prospective cohort studies, another two (10.5%) were cross-sectional studies, and the remaining two were of a mixed cohort study design (10.5%). Sixty-one (75.3%) of the included studies were single-country studies conducted in Northern America, Europe and Asia, five (6.2%) were multi-country, single-continent studies and the remaining 15 (18.5%) were multi-country, multi-continent studies. Overall, the selected studies were conducted in 51 countries distributed in all continents (Table 1).
[Table 1 to be inserted here]
According to the disease stage of the included population, 31 (38.3%) studies were classified as LA HNC, 45 (55.6%) as RM HNC, and the remaining five (6.2%) as Other. The selected studies cumulatively included 9607 LA and/or RM HNC patients. Median patient age ranged from 47 to 78 years across studies (Table 1).
Prevalence of HPV in HNC
The proportion of HPV+ patients over HNC patients enrolled in each study, i.e., HPV prevalence per study, and overall HPV prevalence are presented in Fig. 2 and Table 2. The prevalence of HPV in HNC varied considerably across studies, ranging from 2.9% to 100.0%, with a mean value of 32.6%. To account for variations in sample size of each included study, the pooled HPV prevalence was also calculated across studies and was found to be 25.1%. In the IS (n=62), the prevalence of HPV ranged from 2.9% to 100.0%, with a mean value of 34.5% and a pooled HPV prevalence of 27.1%; while in NIS (n=19) the prevalence of HPV ranged from 3.3% to 47.6%, with a mean value of 26.5% and a pooled HPV prevalence of 19.4%. In a further analysis by disease stage and regardless of study design, prevalence of HPV was examined in the subgroups of patients with LA and RM, as these represent distinct disease phenotypes with different management approaches and survival outcomes. In LA HNC studies (n=31) HPV prevalence ranged from 10.3% to 100.0% (mean 44.7%), with a pooled fraction of 44.0%, while in RM HNC studies (n=45), HPV prevalence ranged from 2.9% to 55.6% (mean 24.3%), with a pooled fraction of 18.6%. Interestingly, in about one sixth of the studies, the prevalence of HPV exceeded 50.0%, indicating that the infection can account for a substantial proportion of HNC in certain patient populations.
[Figure 2 to be inserted here]
Table 2 HPV prevalence in LA and RM HNC and OPC, and OPC fraction, per design and stage
HPV prevalence in HNC
|
|
nstudies
|
Mean
|
Median
|
Range (min, max)
|
NHNC pts enrolled
|
NHPV+ HNC pts
|
Pooled
|
Interventional Studies
|
|
|
|
|
|
|
|
LA
|
26
|
47.6%
|
43.6%
|
10.3% - 100.0%
|
1624
|
812
|
50.0%
|
RM
|
34
|
25.2%
|
23.7%
|
2.9% - 55.6%
|
5484
|
1119
|
20.4%
|
Other
|
2
|
NA
|
NA
|
3.1% - 40.0%
|
42
|
5
|
NA
|
Overall
|
62
|
34.5%
|
30.4%
|
2.9% - 100.0%
|
7150
|
1936
|
27.1%
|
Non-interventional Studies
|
LA
|
5
|
29.5%
|
32.7%
|
16.0% - 37.4%
|
615
|
173
|
28.1%
|
RM
|
11
|
21.4%
|
21.5%
|
3.3% - 44.4%
|
1551
|
189
|
12.2%
|
Other
|
3
|
NA
|
NA
|
30.3% - 47.6%
|
291
|
114
|
NA
|
Overall
|
19
|
26.5%
|
28.6%
|
3.3% - 47.6%
|
2457
|
476
|
19.4%
|
Interventional and non-interventional studies
|
LA
|
31
|
44.7%
|
37.4%
|
10.3% - 100.0%
|
2239
|
985
|
44.0%
|
RM
|
45
|
24.3%
|
23.4%
|
2.9% - 55.6%
|
7035
|
1308
|
18.6%
|
Other
|
5
|
NA
|
NA
|
3.1% - 47.6%
|
333
|
119
|
NA
|
Overall
|
81
|
32.6%
|
29.6%
|
2.9% - 100.0%
|
9607
|
2412
|
25.1%
|
OPC fraction in HNC
|
|
nstudies
|
Mean
|
Median
|
Range (min, max)
|
NHNC pts enrolled
|
NOPC pts enrolled
|
Pooled
|
Interventional Studies
|
|
|
|
|
|
|
|
LA
|
25
|
65.8%
|
67.8%
|
37.9% - 100.0%
|
1585
|
1090
|
68.8%
|
RM
|
27
|
38.3%
|
37.5%
|
4.3% - 75.0%
|
4213
|
1645
|
39.0%
|
Other
|
1
|
NA
|
NA
|
NA
|
10
|
8
|
NA
|
Overall
|
53
|
52.0%
|
50.0%
|
4.3% - 100.0%
|
5808
|
2743
|
47.2%
|
Non-interventional Studies
|
LA
|
5
|
60.4%
|
68.8%
|
45.5% - 71.8%
|
615
|
386
|
62.8%
|
RM
|
8
|
39.2%
|
33.0%
|
23.0% - 67.4%
|
1499
|
609
|
40.6%
|
Other
|
3
|
NA
|
NA
|
NA
|
291
|
166
|
NA
|
Overall
|
16
|
49.4%
|
46.7%
|
23.0% - 76.8%
|
2405
|
1161
|
48.3%
|
Interventional and non-interventional studies
|
LA
|
30
|
64.9%
|
68.1%
|
37.9% - 100.0%
|
2200
|
1476
|
67.1%
|
RM
|
35
|
38.5%
|
37.2%
|
4.3% - 75.0%
|
5712
|
2254
|
39.5%
|
Other
|
4
|
NA
|
NA
|
NA
|
301
|
174
|
NA
|
Overall
|
69
|
51.4%
|
50.0%
|
4.3% - 100.0%
|
8213
|
3904
|
47.5%
|
HPV prevalence in OPC
|
|
nstudies
|
Mean
|
Median
|
Range (min, max)
|
NOPC pts enrolled
|
NHPV+ OPC pts
|
Pooled
|
Interventional Studies
|
|
|
|
|
|
|
|
LA
|
22
|
64.9%
|
70.0%
|
24.0% - 100.0%
|
829
|
557
|
67.2%
|
RM
|
20
|
50.4%
|
52.8%
|
18.9% - 88.9%
|
1436
|
643
|
44.8%
|
Other
|
1
|
ΝΑ
|
NA
|
ΝΑ
|
8
|
4
|
NA
|
Overall
|
43
|
57.8%
|
59.2%
|
18.9% - 100.0%
|
2273
|
1204
|
53.0%
|
Non-interventional Studies
|
LA
|
3
|
62.5%
|
72.0%
|
35.7% - 79.7%
|
98
|
70
|
71.4%
|
RM
|
5
|
29.1%
|
21.4%
|
14.3% - 71.4%
|
292
|
60
|
20.5%
|
Other
|
1
|
ΝΑ
|
NA
|
ΝΑ
|
51
|
42
|
NA
|
Overall
|
9
|
46.2%
|
35.7%
|
14.3% - 82.4%
|
441
|
172
|
39.0%
|
Interventional and non-interventional studies
|
LA
|
25
|
64.6%
|
70.0%
|
24.0% - 100.0%
|
927
|
627
|
67.6%
|
RM
|
25
|
46.1%
|
44.4%
|
14.3% - 88.9%
|
1728
|
703
|
40.7%
|
Other
|
2
|
ΝΑ
|
NA
|
ΝΑ
|
59
|
46
|
NA
|
Overall
|
52
|
55.8%
|
57.9%
|
14.3% - 100.0%
|
2714
|
1376
|
50.7%
|
Abbreviations: HNC, head and neck cancer; HPV, human papilloma virus; LA, locally-regionally advanced; Npts, number of patients; n, number of studies; NA, not applicable; OPC, oropharyngeal cancer; RM, recurrent/metastatic.
Prevalence of HPV in OPC
Given the increasing incidence of OPC reported previously, and the proposed contribution of HPV to this increase, the proportion of HPV+ patients was also assessed among the subgroup of patients with LA and/or RM OPC in the studies of the evidence synthesis. The proportion of patients with OPC among those with LA and/or RM HNC, referred to as the OPC fraction, is presented in Fig. 2 and summarized in Table 2. The OPC fraction among studies with available HN sub-site proportions (n=69) ranged from 4.3% to 100.0%, with a mean of 51.4%. Based on pooled data, of the 8213 LA and/or RM HNC patients, 3904 had OPC, resulting in a pooled fraction of 47.5%. The mean (and pooled) fractions in IS (n=53; range 4.3% to 100.0%) and NIS (n=16; range 23.0% to 76.8%) were 52.0% (47.2%) and 49.4% (48.3%), respectively. Upon analysis by disease stage, the mean (and pooled) OPC fraction was 64.9% (67.1%) in LA HNC studies (n=30; range 37.9% to 100.0%) and 38.5% (39.5%) in RM HNC studies (n=35; ranging from 4.3% to 75.0%) (Table 2).
HPV prevalence in LA and/or RM OPC was available for 52 studies and ranged from 14.3% to 100.0%, with a mean value of 55.8% and a pooled fraction of 50.7%. HPV prevalence in LA and/or RM OPC ranged from 18.9% to 100.0% in IS, and from 14.3% to 82.4% in NIS with available data, with respective mean (and pooled) rates of 57.8% (53.0%) and 46.2% (39.0%). Upon analysis by disease stage, the mean (and pooled) HPV prevalence in LA OPC studies was 64.6% (67.6%), ranging from 24.0% to 100.0%, while in RM OPC studies it was 46.1% (40.7%), ranging from 14.3% to 88.9% (Table 2).
Geographic distribution of HPV prevalence
To gain insight into the availability of published data on the prevalence of HPV across geographical regions, as well as to qualitatively assess potential variations among countries or regions, the geographic distribution of HPV prevalence was addressed as an exploratory objective. Of the 53 countries where the studies included in the analysis of the present review were conducted, 29 were located in Europe, 13 in Asia (including Taiwan and Hong Kong and Taiwan as territories of China), 5 in Southern America, 2 countries each in Northern America and Africa, and 1 country each in Central America and Oceania. Τhe following were included in more than ten studies each: United States of America (USA) (44 studies), Germany (16), France (15), Spain (13), Italy (12), Belgium (11), Canada (11), and the United Kingdom (11) (Table 1). Thus, although studies with published data on HPV prevalence in LA and RM HNC through the last decade display a wide geographic distribution, several geographic regions are underrepresented in the literature and further studies would be needed to more accurately capture the global epidemiological picture.
The prevalence of HPV in LA and RM HNC and OPC is summarized per geographical region in Table 3 and in Additional file 4, while it is also presented per disease stage in Additional file 5. Based on the geographic regions included, studies can be broadly divided into those conducted in a single continent and those conducted in multiple continents. In single-continent HNC studies conducted in Northern America (n=34), the prevalence of HPV ranged from 8.3% to 100.0%; in Europe (n=29) from 3.1% to 75.9%; in Eastern Asia (n=3) from 10.3% to 33.3%. The mean (and pooled) prevalence of HPV among single-continent studies conducted in Northern America was 46.0% (42.1%), followed by 24.7% (25.3%) in Europe, and 20.1% (15.7%) in Eastern Asia. Studies conducted in Europe were also grouped into those conducted in Northern Europe, Southern Europe, Western Europe, or Multiple European regions (including Western, Central/Eastern, and Southern Europe) based on data availability. The respective mean (and pooled) HPV prevalence was 31.9% (63.1%), 23.2% (26.4%), 24.3% (23.5%), and 17.2% (9.4%). In studies conducted in multiple continents (n=15) the prevalence of HPV ranged from 2.9% to 30.4%, and the mean (and pooled) prevalence of HPV was 19.8% (18.4%).
Table 3 HPV prevalence in LA and RM HNC and OPC, and OPC fraction, per geographic region
HPV prevalence in HNC
|
|
nstudies
|
Mean
|
Median
|
Range (min, max)
|
NHNC pts enrolled
|
NHPV+ HNC pts
|
Pooled
|
Northern America
|
34
|
46.0%
|
43.6%
|
8.3% - 100.0%
|
1923
|
809
|
42.1%
|
Europe
|
29
|
24.7%
|
22.7%
|
3.1% - 75.9%
|
2804
|
710
|
25.3%
|
Northern Europe
|
5
|
31.9%
|
20.7%
|
11.5% - 75.9%
|
377
|
238
|
63.1%
|
Southern Europe
|
4
|
23.2%
|
26.6%
|
3.3% - 36.4%
|
284
|
75
|
26.4%
|
Western Europe
|
17
|
24.3%
|
25.0%
|
3.1% - 47.6%
|
1381
|
325
|
23.5%
|
Multiple European regions†
|
3
|
17.2%
|
13.3%
|
6.8% - 31.6%
|
762
|
72
|
9.4%
|
Eastern Asia
|
3
|
20.1%
|
16.7%
|
10.3% - 33.3%
|
121
|
19
|
15.7%
|
Multiple continents
|
15
|
19.8%
|
21.6%
|
2.9% - 30.4%
|
4759
|
874
|
18.4%
|
OPC fraction in HNC
|
|
nstudies
|
Mean
|
Median
|
Range (min, max)
|
NHNC pts enrolled
|
NOPC pts enrolled
|
Pooled
|
Northern America
|
31
|
56.9%
|
57.8%
|
4.3% - 100.0%
|
1797
|
1051
|
58.5%
|
Europe
|
24
|
53.7%
|
51.9%
|
15.4% - 85.2%
|
2462
|
1371
|
55.7%
|
Northern Europe
|
5
|
41.7%
|
33.3%
|
15.4% - 85.2%
|
377
|
273
|
72.4%
|
Southern Europe
|
4
|
47.8%
|
45.6%
|
23.0% - 76.9%
|
284
|
128
|
45.1%
|
Western Europe
|
13
|
60.3%
|
60.2%
|
37.5% - 80.0%
|
1279
|
777
|
60.8%
|
Multiple European regions†
|
2
|
52.1%
|
52.1%
|
35.8% - 68.4%
|
522
|
193
|
37.0%
|
Eastern Asia
|
2
|
33.0%
|
33.0%
|
22.9% - 43.1%
|
106
|
36
|
34.0%
|
Multiple continents
|
12
|
36.0%
|
36.5%
|
13.8% - 58.5%
|
3848
|
1446
|
37.6%
|
HPV prevalence in OPC
|
|
nstudies
|
Mean
|
Median
|
Range (min, max)
|
NOPC pts enrolled
|
NHPV+ OPC pts
|
Pooled
|
Northern America
|
25
|
70.0%
|
75.0%
|
21.4% - 100.0%
|
577
|
423
|
73.3%
|
Europe
|
17
|
44.1%
|
44.4%
|
14.3% - 89.1%
|
829
|
426
|
51.4%
|
Northern Europe
|
4
|
61.4%
|
65.7%
|
25.0% - 89.1%
|
264
|
230
|
87.1%
|
Southern Europe
|
4
|
48.2%
|
49.3%
|
14.3% - 79.7%
|
128
|
75
|
58.6%
|
Western Europe
|
7
|
35.3%
|
35.7%
|
20.0% - 50.0%
|
244
|
81
|
33.2%
|
Multiple European regions†
|
2
|
32.5%
|
32.5%
|
18.9% - 46.2%
|
193
|
40
|
20.7%
|
Eastern Asia
|
1
|
NA
|
NA
|
NA
|
25
|
6
|
24.0%
|
Multiple continents
|
9
|
41.6%
|
38.9%
|
15.8% - 59.2%
|
1283
|
521
|
40.6%
|
Abbreviations: HNC, head and neck cancer; HPV, human papilloma virus; Npts, number of patients; n, number of studies; NA, not applicable; OPC, oropharyngeal cancer.
† The category of Multiple European regions includes multi-country studies conducted in Europe. These studies were conducted in Western, Central/Eastern, and Southern Europe for HNC, and Southern and Western Europe for OPC.
Among studies with available HN sub-site proportions (regardless of HPV status), mean (and pooled) OPC fraction was 56.9% (58.5%), 53.7% (55.7%), 33.0% (34.0%), and 36.0% (37.6%) in studies conducted in Northern America (n=31), Europe (n=24), Eastern Asia (n=2) and multiple continents (n=12), respectively (Table 3). Moreover, based on the proportion of HPV+ OPC patients in studies with available data, the mean (and pooled) prevalence of HPV in LA and RM OPC was 70.0% (73.3%) in studies conducted in Northern America (n=25); 44.1% (51.4%) in studies conducted in Europe (n=17), and 41.6% (40.6%) in studies conducted in multiple continents (n=9). In the only single-country study conducted in Eastern Asia, the prevalence of HPV in LA and RM OPC was 24.0% (Table 3). Within Europe, the mean (and pooled) prevalence of HPV in LA and RM OPC was 61.4% (87.1%) in Northern Europe, 48.2% (58.6%) in Southern Europe, 35.3% (33.2%) in Western Europe, and 32.5% (20.7%) in multiple European regions (including Southern and Western Europe). Taken together, the above data illustrate high rates of HPV prevalence in LA and RM HNC and OPC across different geographical regions.
HPV detection techniques
In the absence of HPV diagnostic tests with regulatory approval for HNC over the examined period, and given that HPV testing is generally recommended for all newly diagnosed OPSCC but is not warranted for the other HNC types, the present review aimed to capture HPV detection techniques utilized in the included studies. HPV detection techniques are retrieved and analyzed as reported by the authors in the publications. Information on reported HPV detection assays across the included HNC studies are presented in Additional file 6. In total, HPV status was assessed in any HN anatomical site in 37 studies (45.7%), in OPX only in 38 studies (46.9%) while six studies (7.4%) did not provide information on the site examined. With respect to specific methodologies, of the 81 studies, 47 (58.0%) reported using a p16INK4a-based method, two studies (2.5%) employed quantitative Reverse Transcription-Polymerase Chain Reaction (qRT-PCR), two studies (2.5%) employed in situ hybridization (ISH), while in one study each the detection method was referred to as DNA testing, quantitative Reverse Transcription-PCR (qRT-PCR), and immunohistochemistry (IHC). Seven studies (8.6%) reported using multiple detection techniques to determine HPV status at a cohort level, even though at a patient level HPV status could also have been derived solely based on a single technique. For the remaining 20 (24.7%) studies the authors did not provide any relevant information. HPV detection methods are also presented for IS and NIS, by disease stage, and site examined in Additional file 6. Irrespective of grouping, “p16-based” detection methodologies were the most frequently reported across studies.
HPV prevalence in OPC using solely a p16-based method
Considering that p16 overexpression is generally used as a surrogate marker for the presence of HPV in OPSCC and the recommendation for p16 testing in OPSCC clinical management [38, 39], a supplementary analysis was performed by isolating the studies reporting solely a p16-based method for HPV testing and having available results in OPC. In total, 30 studies were included in this analysis (26 IS and 4 NIS; 16 LA and 14 RM), with prevalence of HPV ranging from 15.8% to 100.0% and a mean (and pooled) HPV prevalence of 57.2% (52.6%) (Additional file 7), further supporting the main outcomes of this evidence synthesis.
Distribution of HPV prevalence by number of enrolled HNC patients
As a means to evaluate the potential effect of variations across individual sample sizes on the primary outcome of overall prevalence of HPV in LA and RM HNC, the prevalence of HPV reported for each included study was plotted against the respective sample size (Fig. 3). No obvious asymmetry was observed around the calculated overall mean HPV prevalence. Based on this distribution, no apparent bias in the estimation of the study primary outcome arising from sample size can be inferred.
[Figure 3 to be inserted here]