Head and neck squamous cell carcinoma (HNSCC) typically develops in the sixth to seventh decade of life. Since Byers identified this subset of patients in 1975, clinicians have become increasingly aware of patients who develop HNSCC at a young age, variably defined as age 30 years and younger, 40 years and younger, or 50 years and younger.  These patients may represent a distinct cohort with different risk factors and disease behavior.
This article addresses this cohort, defined here as young adults, age 20-45, with HNSCC. As discussed below, these cancers tend to occur in the oral cavity and oropharynx, rather than other head and neck subsites. This article primarily addresses squamous cell carcinoma (SCC), but other histology is noted when appropriate. This article does not address pediatric head and neck tumors, and the reader is referred to other chapters for this topic. See the image below.
The incidence of HNSCC in young patients is approximately 1-8% of all head and neck cancers, based on modern reports from the United States, Canada, Great Britain, Spain, Scandinavia, India, and Japan. [2, 3, 4, 5, 6, 7, 8, 9, 10, 11] Current evidence suggests that this incidence may be increasing. [12, 13]
From 1973-2001, Shiboski et al demonstrated an increase of approximately 1-4% in the incidence of oral cavity and pharynx cancer among young whites based on subsite.  Similarly, Schantz and Yu noted a 60% increase in the incidence of tongue cancer in young patients from 1973-1997, with no increase in incidence for other age groups.  Both studies used National Cancer Institute Surveillance, Epidemiology, and End Results (SEER) data for their analyses. The etiology of this increasing trend is unclear, although marijuana use and human papillomavirus (HPV) infection are possible explanations. [4, 7, 3]
Although HNSCC generally remains more common in males, even among young patients, some studies have reported a higher relative incidence in females. [8, 15, 11, 16] Furthermore, one report demonstrated a reversal of the typical male-to-female ratio in favor of women within the 35-year-old or younger age group. 
Additionally, these cohorts of young women lack the typical associated risk factors of alcohol, tobacco, and betel nut exposure and may represent a unique subset even within young HNSCC patients. [10, 16, 17, 1]
Funk proposed the following 3 groupings of young patients with head and neck cancer: 
Group I is composed of young women (< 35 years) with few to no risk factors and aggressive squamous cell carcinoma (SCC)
Group II are typically young males (< 40 years) with heavy risk factors and disease typical of older patients
Group III has a slight male predominance but fewer risk factors and relatively well-differentiated disease.
In Funk’s review of oral cancer in the National Cancer Data Base (NCDB) between 1985 and 1996, the predominant group was Group III.  If this categorization is valid, the group of patients predominating any given study may change the significance of gender and other factors.
Among the few studies that address race, conflicting data are found in young patients with HNSCC. Shiboski et al found an overall higher percentage of oral cavity and pharynx cancer among young whites compared with young African Americans from a 1973-2001 SEER database analysis and found the incidence of these cancers increased in that time period only among young whites. 
However, Kolker et al, looking at statistics from metropolitan Detroit, a predominantly African American area of Michigan, found a higher incidence of oral and pharyngeal cancer among African Americans compared with whites in the 50 and under age group.  Similarly, Slotman et al found a higher incidence of HNSCC in African Americans younger than 45 years in 2 disparate locations.  The first was in East Orange, Virginia, where 21% of the 847 patients studied from 1968-1980 were African American. The second was the New Jersey Medical School, where 74% of the 219 patients studied from 1970-1980 were African American.
Both sites found a trend toward younger age at presentation for African Americans. In the first site, 13% of African Americans presented before age 45, compared with 3% of whites; in the second site, 15.3% of African Americans presented before age 45, compared with 2% of whites. Slotman and colleagues also noted a lower 5-year survival rate for African Americans in all age groups.
No data exist in the above studies to explain the racial differences in age at presentation. Slotman and colleagues suggested that their African American patients may have started using tobacco and alcohol at a younger age, but the authors did not provide comprehensive data to support this theory.  Clearly, except for the SEER data, the above studies are small and subject to the populations they survey, which may bias their findings.
HNSCC in the young patient tends to occur in the oral cavity and oropharynx.  In the oral cavity, the oral tongue is the most common subsite in the young in most series; [8, 14, 20, 16, 11, 21] the floor of the mouth has been reported as a less common subsite in the young patient compared with older patients. 
The incidence of oral tongue and tonsil cancer seems to be increasing among younger patients. [22, 14, 4, 23] The association of oropharyngeal cancer with HPV infection (see Etiology and Risk Factors) lends circumstantial evidence to that virus’s role in the etiology of HNSCC in the young adult. Because HPV’s role in the etiology of oral cancer is controversial, HPV carcinogenesis is not yet a convincing explanation for the increase in both oral cavity and oropharyngeal tumors.
Within the oral cavity, certain non–SCC histologies also account for a substantial proportion of cancers in the young patient. In particular, adenocarcinoma and Kaposi sarcoma tend to occur on the palate, making the palate another oral cavity subsite that is more common in young patients compared with older patients. [16, 24]
Many of the studies comparing young patients who have HNSCC with old patients with HNSCC use tumor stage as a matching criteria for creating the older, "control" cohort, thus limiting the data comparing stage at presentation. Only those studies looking at whole populations can reliably comment on differences in stage at presentation. Schantz and Yu reviewed the 1973-1997 SEER database and found younger patients to be more likely to present with localized disease than older patients.  Similarly, Funk’s review of the 1985-1996 NCDB demonstrated younger patients to present at an earlier stage across all types of histology, and a statistically significant higher proportion of stage I disease among young patients when analyzed only for squamous cell carcinoma (SCC). 
Contrary to these findings, in small single-institution studies, Verschuur and colleagues and Veness and colleagues both found a higher rate of nodal metastases at presentation in younger patients. [3, 26] However, Veness and colleagues also found older patients to be more likely to present with a higher T stage compared with younger patients; this latter finding was echoed by Sasaki and colleagues in their single-institution series.  In summary, although a propensity for increased nodal metastases among some young patients may exist, this has yet to be clearly defined, and evidence from the largest databases suggests young patients present at similar or earlier stages than older patients.
In Byers’s report from 1975, his young patients (< 30 years old) with oral tongue squamous cell carcinoma (SCC) had a high percentage (almost 50%) of high-grade histology.  Four of those patients were female and 7 were male. Since that publication, histology has been addressed specifically by a handful of studies that contradict the assertion that younger patients tend to have more aggressive histology. Atula and colleagues reviewed 34 Finnish patients younger than 40 years with squamous cell carcinoma (SCC) of the tongue and found the vast majority (70%) to have well-differentiated tumors.  No older population was reported for comparison. Similarly, Sasaki and colleagues found 66% of young tumors (patients < 40 years) to be well-differentiated, compared with only 33% in their older cohort. 
However, in a publication focused specifically on margin assessment in oral cavity tumors, Spiro and colleagues incidentally found young patients, defined as younger than age 50 years, to make up a higher percentage of high-grade lesions (38%) than low-grade lesions (17%).  This paper scored tumors based on infiltration at the margins, with higher-grade lesions demonstrating infiltrative cords or nests rather than the "pushing" border of lower-grade lesions. In Spiro and colleagues’ series, those patients with higher-grade lesions tended to present with nodal metastases and distant disease but did not have a higher rate of locoregional recurrence compared with lower-grade lesions. Unfortunately, a subset analysis for young patients was not performed for these parameters, so one cannot comment definitively on stage at presentation or recurrence from this paper.
Using Funk’s grouping, Byers’s findings may be explained by involving primarily the unique subset of young patients with aggressive disease (Group I). Spiro’s findings would not fall into that category, and indeed Spiro’s definition of young was considerably older than Byers’ definition. Spiro also had many older patients as well as younger patients with aggressive histology. A bimodal distribution of aggressive histology (the youngest patients and older patients having more aggressive histology) is possible, but this cannot be proven from the available data.
Finally, note again that not all oral cavity or head and neck tumors in the young are squamous cell carcinoma (SCC). Funk found a higher percentage of non–SCC histology, particularly adenocarcinoma and Kaposi sarcoma, in patients younger than 35 years compared with their older counterparts. In that series, approximately 48% of oral cavity cancers in patients younger than 35 years were SCC, versus 88% in patients 36-65 years and 91% of patients older than 65 years. 
Etiology and Risk Factors
Tobacco and alcohol
In young patients, studies have found a variable and sometimes absent relationship with these traditional risk factors. Several small studies have fallen on both sides of this debate, with some finding a lower rate of tobacco and alcohol use among young head and neck squamous cell carcinoma (HNSCC) patients compared with older HNSCC patients; [29, 29, 7, 11, 30] however, others found no difference in use. [31, 26, 32]
In a review of risk factors in 116 patients from the south east of England, Llewellyn et al found equal and substantial exposure to tobacco and alcohol in young patients with oral SCC and a control group of patients without cancer (both groups were composed of subjects younger than 45 years).  In their analysis, tobacco consumption for greater than 21 years resulted in a significantly elevated risk of oral cancer. Note that patients starting tobacco use in their teenage years reach the 21-year mark prior to their 40th birthday.
In a smaller, more recent series of 56 patients with oral cavity cancer and controls matched for age, Llewellyn and colleagues found an increased risk of cancer in males who started smoking before their 16th birthday, reinforcing that tobacco use can still be a risk factor in the young.  Similarly, Lipkin and colleagues found a very high rate of heavy alcohol and tobacco use in an analysis of 39 adults with HNSCC age 40 or younger from the Baylor College of Medicine. 
The use of smokeless tobacco has been linked to oral cavity cancer. In an analysis of the SEER database from 1973-1984, Davis and Severson found tongue cancer to increase by 1.8 fold in the 30-39 age group.  They juxtaposed their findings to other data that demonstrated an increase in the use of smokeless tobacco among the young. However, as the SEER database did not contain specific data on habits of smokeless tobacco use, the association is conjectural.
Interestingly, a third series on young patients with oral squamous cell carcinoma (SCC) by Lewellyn et al highlighted a caveat regarding the connection between tobacco and cancer: an absence of tobacco use doesn’t mean an absence of cancer. They reported that among young patients with cancer, a lack of tobacco use was associated with a delay in seeking medical care for their cancer-related symptoms. 
Thus, young patients with no risk factors (and their physicians) may not suspect cancer despite worrisome signs and symptoms. Clearly, subsets of young patients who present without these risk factors (Funk’s Groups I and III) exist; in fact, 38% of the young women and 18% of the young males in Llewellyn’s larger series had no risk factors.  Such patients have been the subject of searches for additional etiologies specific to or present in young patients with HNSCC, as addressed below.
Marijuana use and its association with younger patients has led to suggestions that this may be an unreported risk factor in studies that have found fewer risk factors in younger patients. In the only studies to specifically address marijuana use in young patients, 2 of Llewellyn et al’s series both reported similar use of marijuana among young patients with oral SCC and young controls. [17, 5]
Zhang and colleagues at Memorial Sloan Kettering Cancer Center looked at the incidence of marijuana use among patients with pathologically confirmed squamous cell carcinoma (all head and neck sites) compared with age-matched and sex-matched controls culled from their blood bank records.  Most marijuana smokers from both groups were under the age of 55. An increased odds ratio for risk of squamous cell carcinoma among marijuana smokers was found, but, as the authors point out, whether their data was biased by the controls being blood donors (and possibly less likely to be users of illegal drugs) was unclear.
In an age- and sex-matched, case control analysis of HPV-positive and HPV-negative HNSCC at Johns Hopkins, Gillison et al found an association of HPV positivity and marijuana use that increased in strength with increasing cumulative marijuana use.  In contrast, the HPV-negative group was more strongly associated with traditional HNSCC risk factors (tobacco, alcohol, and poor oral hygiene). No specific analysis of marijuana use and age with respect to HNSCC in this study was found, but the study does provide further evidence for a subgroup of patients who lack traditional risk factors yet share the risk factors of HPV positivity and marijuana use that many have linked to younger patients. Also, the observed cohorts of younger patients with head and neck cancer that have been reported in previous decades are possibly a result of a wave of generations in which marijuana use had become more commonplace during the teenage years.
In 2 of Llewellyn and colleagues’ series, a significant reduction in risk was found in subjects who reported consumption of 3 or more portions of fresh fruit or vegetables per day. [17, 5] Generally, a diet high in fruits and vegetables is inversely correlated with a risk of oral cancer,  and, based on Lewellyn’s studies, this can also be applied to young patients.
Perhaps the most widely studied virus in the head and neck cancer literature in recent years is HPV, a virus initially linked to cervical carcinogenesis that has now gained interest for its connection to cancer of the oropharynx, particularly the lingual and palatine tonsils. [36, 37] The increasing incidence of tongue and tonsil cancer among young patients has led some authors to suggest that HPV may be responsible for this trend; [4, 14] although the connection between oral (versus oropharyngeal) cancer and HPV is controversial. [38, 39]
In a direct comparison between oral cavity, oropharynx, and larynx squamous cell carcinoma (SCC) samples from young (< 50 years) and old (>50 years) patients, Sisk and colleagues found no significant difference by age in the rates of HPV positivity (50% vs 44%, respectively.  The authors acknowledge that theirs was a small study that does not allow definitive conclusions on this matter. Similarly, Koch and colleagues did not find an association of HPV positivity with age in 305 patients with HNSCC. 
Gillison and colleagues, however, in a case control analysis looking at HPV-16 status of 240 patients with HNSCC at Johns Hopkins, found a higher proportion of young patients (< 50 years) in the HPV-16–positive group than the HPV-16–negative group (33% vs 17%, respectively).  Additionally, they found a strong association between HPV-16 positivity and oropharyngeal and lingual or palatine tonsil primary sites.
In a multi-institutional, prospective phase II trial of chemoradiation for advanced HNSCC (ECOG 2399) that included analysis of HPV status in the primary biopsy, the mean age of the 38 patients with HPV positive tumors was lower than the 58 with HPV-negative tumors (56 versus 60, respectively), but the difference was not significant.  However, this study did find a significantly better response and survival in the HPV-positive group, even when adjusted for age, tumor stage, and performance status.
Sisk likewise found HPV positivity to be linked to a better overall prognosis,  and this has been seen in several other studies as well,  although none of these studies specifically addressed young patients. Interestingly, improved survival was also seen with increasing copy number of HPV-16 in 35 patients with tonsil carcinoma in one study, suggesting that the connection between HPV status and response to treatment may be quite strong. 
The connection between HPV and oropharyngeal cancer combined with the current evidence suggesting a better prognosis with HPV-related tumors may be one explanation for the markedly different subsets of young patients with head and neck cancer. If tongue cancer, for instance, is not related to HPV, but rather a more sinister, as of yet unidentified risk factor, this would explain the reports by Byers and others that describe young women with aggressive and difficult-to-treat disease (Funk’s Group I).
The Gillison and colleagues’ case control analysis did not address prognosis, but they did find the HPV-16–positive group to be significantly less likely to have the typical risk factors for HNSCC (tobacco, alcohol, poor dental hygiene).  Funk’s Group III patients fit into this description. Clearly, more research is needed into the role of HPV in HNSCC and its connection to treatment response, as well as into possible chemopreventive strategies (like those developed for cervical cancer).
Human immunodeficiency virus
Infection with the human immunodeficiency virus (HIV) and progression to AIDS is positively correlated with malignancies of the upper aerodigestive tract, particularly Kaposi sarcoma and non-Hodgkin lymphoma and, to a lesser extent, squamous cell carcinoma (SCC).  Funk and colleagues reported a very high incidence of Kaposi sarcoma in young African American males between 1985 and 1996, a time when the AIDS epidemic heavily affected that demographic.  Other series have not addressed (even indirectly) immunodeficiency due to HIV or other causes in young patients.
Some genetic component to cancer development in young patients is likely, particularly in those patients with no recognized risk factors. These patients have been shown to have increased DNA fragility, which may make them more likely to develop genetic abnormalities.  However, studies examining specific genetic alterations in HNSCC as a function of patient age, including mutations in p53, p21, Rb, and MDM2, [27, 46] and microsatellite instability,  have failed to find an increase in these abnormalities among young patients.
Likewise, Koch and colleagues found tumors of nonsmokers with HNSCC to have fewer genetic abnormalities than those of their smoking counterparts, leading them to conclude that the genetic alterations in tumors from nonsmoking patients remain undiscovered.  Llewellyn and colleagues’ larger series reported a positive family history of cancer in 75% of female and 59% of male subjects younger than age 45 with HNSCC, suggesting genetics, immunology, or some common environmental exposure may play a role in the development of HNSCC in the young. 
A rare inherited cancer syndrome associated with HNSCC is Fanconi anemia (FA), an autosomal recessive syndrome caused by defects in DNA repair. FA carries a high risk of development of malignancy at a young age, with a median age of presentation of 31, and a cumulative incidence of HNSCC of 14% by age 40. [48, 49] Patients with FA and HNSCC are more likely to be female (2:1), with very few reporting tobacco use. Oral cavity is the most common site and the outcome is poor, with a 63% rate of second primaries and a 2-year overall survival of 49%. 
No specific additional workup is required for young patients with head and neck squamous cell carcinoma (HNSCC) compared with their older counterparts. However, a young patient presenting with no traditional risk factors may warrant at least a careful history to examine for other possible etiologies, including a familial syndrome or evidence of immunodeficiency. If these etiologies are uncovered, appropriate additional workup should be carried out. In centers with active protocols examining the role of human papillomavirus (HPV) or other genetic abnormalities, eligibility for these studies should be assessed and participation offered. As of yet, no role for HPV serotyping or genetic testing in young patients has been standardized.
Please see sections on specific head and neck subsites for workup of malignancies in these areas.
Treatment & Management
A study by Murphy et al, using information from the National Cancer Data Base, indicated that for patients with head and neck cancer, extending the time to treatment initiation (TTI) beyond 46-52 days has a negative effect on overall survival (OS). The study, which involved 51,655 patients, found that when TTI was 46-52 days or less, median OS was 71.9 months, while median OS for a TTI of 53-67 days was 61 months and median OS for a TTI of longer than 67 days was 46.6 months. [50, 51]
Chemotherapy and radiation therapy
No studies have specifically examined different types or doses of chemotherapy or radiation therapy as a function of age. Most studies using chemotherapy as a modality for head and neck squamous cell carcinoma (HNSCC) fail to stratify age in looking at outcome. Currently, available studies looking at outcomes with regards to patient age give little to no data about specific chemotherapeutic agents or radiation techniques and dosing. Given that less than 2% of HNSCC patients are enrolled in clinical trials nationwide, only future enrollment of all patients into clinical trials can give us knowledge about the role of modality and outcome regarding patient age.
Radiation is used as single modality therapy for early stage disease (stage I-II) and in multimodality therapy for advanced disease (stage III-IV). However, radiation of young patients has specific implications given their potential lifespan. As patients survive longer after radiation treatment for head and neck cancer, the long-term consequences of this treatment become more significant. Difficulties with xerostomia, fibrosis, and swallowing are significant quality of life issues in long-term survivors of head and neck irradiation. [52, 53]
Radiation-induced malignancies, such as sarcoma or thyroid carcinoma, although uncommon, are also a concern in the long-term follow-up of irradiated patients. Young HNSCC patients successfully treated with radiation have longer posttreatment lives in which to potentially develop these malignancies, and this has prompted some authors to suggest surgery as the primary modality for treatment of the young patient. 
Conversely, recent evidence suggests that HPV-positive tumors may be more responsive to organ preservation therapy than HPV-negative tumors.  This may lead to organ preservation therapy being the preferred modality for both young and old patients who are HPV positive.
No studies specifically address the extent of surgical resection in young patients. As discussed in Histology (in the Introduction section) and Outcome and Prognosis, most young patients with HNSCC do not have an inherently more aggressive disease, and, therefore, do not require a departure from standard treatment for any given stage. However, by the same token, efforts should not be made to perform a more limited or less complete resection simply because of a patient’s young age. Further research may reveal if young women with aggressive disease (Funk’s Group I) warrant more aggressive initial treatment, even when presenting with early disease (stage I-II). 
Outcome and Prognosis
The vast majority of publications on young patients with head and neck squamous cell carcinoma (HNSCC) address outcome, and yet debate in the literature continues as to whether age at presentation has any effect on prognosis. Identified papers addressing prognosis in young patients with HNSCC are summarized in Table I below. Many of the studies are smaller, retrospective, matched studies and case series that suffer limitations of this methodology, as well as other methodological issues.  Only one study matched patients by treatment and found no difference in survival among 31 patients younger than 40 years and 62 matched controls. 
Table I. HNSCC Papers (Open Table in a new window)
|Authors||Year||Site(s) [a]||Age||# Young||# Control||Method [b]||Prognosis for Young [c]|
|Venables and Craft||1967||T||< 30||13||none||Case Series||Worse|
|Byers||1975||T||< 30||11||none||Case Series||Worse|
|Amsterdam et al||1982||OC||< 35||12||none||Case Series||Worse|
|McGregor et al||1983||OC||< 40||27||none||Case Series||Better|
|Newman et al||1983||T||< 30||13||none||Case Series||Similar|
|Mendez et al||1985||A||<=40||63||none||Case Series||Similar|
|Son and Kapp||1985||OC, OP||< 40||27||none||Case Series||Worse|
|McGregor and Rennie||1987||OC||< 40||13||none||Case Series||Worse|
|Benninger et al||1988||A||<=40||41||none||Case Series||Worse|
|Cusamano et al||1988||CO, OP||<=40||23||none||Case Series||Worse|
|Tsukuda et al||1993||A||< 40||48||none||Case Series||Better|
|Sarkaria and Harari||1994||T||< 40||6||none||Case Series||Worse|
|Atula et al||1996||T||< 40||34||none||Case Series||Similar|
|Martin-Granizo et al||1997||OC||< 40||24||none||Case Series||Similar|
|Pitman et al||2000||T||< 40||28||none||Case Series||Similar|
|Iype et al||2001||OC||< 35||264||none||Case Series||Similar|
|Vermund et al||1982||T||< 40||16||384||Institutional Series||Better|
|Von Doersten et al||1995||A||< 40||23||122||Institutional Series||Similar|
|Siegelmann-Danieli et al||1998||T||<=45||30||57||Institutional Series||Similar|
|Veness et al||2003||T||<=40||22||142||Institutional Series||Similar|
|Lipkin et al||1985||OC, OP, L||<=40||39||39||Matched Control (sx, si, st)||Similar|
|Schantz et al||1988||A||<=40||83||83||Matched Control (sx, st, si, y)||Worse|
|Kuriakose et al||1992||OC||< 35||37||37||Matched Control (random sample)||Similar|
|Friedlander et al||1998||T||< 40||36||36||Matched Control (sx, st, y)||Similar|
|Verschuur et al||1999||A||< 40||185||185||Matched Control (sx, si, y)||Better|
|Vargas et al||2000||T||< 40||17||17||Matched Control (st)||Worse|
|Pytynia et al||2004||A||< 40||31||62||Matched Control (sx, r, si, st, tx)||Similar|
|Sasaki et al||2005||OC||< 40||35||110||Matched Control (random sample)||Similar|
|Garavello et al||2007||T||<=40||46||92||Matched Control (sx, si, st)||Worse|
|Lee et al||2007||T||< 45||20||20||Matched Control (sx, st)||Better|
|Ho et al||2008||OC||< 45||28||56||Matched Control (sx, si, st)||Better|
|Lacy et al||2000||A||<=40||40||990||Database Review (Washington University Cancer Registry)||Better|
|Davidson et al||2001||T||< 40||N/A [d]||N/A [d]||Database Review (SEER 1988-1993)||Better|
|Annertz et al||2002||T||< 40||276||4748||Database Review (Scandanavian Cancer Registry)||Better|
|Funk et al||2002||OC||< 35||2,148||56776||Database Review (National Cancer Database)||Better|
|Schantz and Yu||2002||A||< 40||3339||60070||Database Review (SEER 1973-1997)||Better|
|Kolker et al||2007||PC, OP||< 50||1343||9401||Database Review (Michigan State Cancer Registry)||Worse|
Table I: Articles addressing prognosis of young patients with head and neck cancer were identified from PubMed searches and from reference lists of primary and review articles.
[a] Site abbreviations are as follows: T=tongue, OC=oral cavity, OP=Oropharynx, L=larynx, A=all head and neck sites.
[b] Matching criteria is as follows: sx=sex, si, site, st=stage, r=race, tx=treatment, y=year at presentation; SEER=Surveillance, Epidemiology, and End Results program.
[c] Relative prognosis of young patients compared with older patients reported here is what was reported by the authors of the identified papers. For all of the case series listed, authors made comparisons to historic "controls" to assign relative prognosis.
[d] Specific numbers are not listed in the paper.
Additionally, remember that young patients are not necessarily free of other medical problems. Singh and colleagues found the presence of advanced comorbidity to diminish disease-free interval and tumor specific survival in an analysis of 70 patients with HNSCC younger than age 45.  This remains another element that is not typically accounted for in matched studies.
The larger database studies may be more reliable. Funk and colleagues analyzed the 1985-1996 NCDB, finding an increased 5-year survival rate for patients with HNSCC age 35 or younger compared with patients age 36-65, and for patients age 36-65 compared with patients 65 years and older across all stages, but these differences were only significant for stage I disease.  Shiboski and colleagues’ analysis of the 1973-2001 SEER database revealed an overall increased 5-year survival for patients with HNSCC who were younger than 45 years.  An analysis of Scandinavian cancer registries by Annertz and colleagues similarly found increased 5-year survival among patients younger than 40 years. 
Lacy and colleagues reviewed the cancer database of the Washington University School of Medicine from 1980-1991.  Forty patients aged 40 years or younger were identified out of 1030 patients with 5-year survival data available. Young patients had a significantly better 5-year survival than the middle-aged (41-64 y) or old (≥ 65 y) age groups (65% vs 52% and 35%, respectively). Young patients also had a lower recurrence rate and a lower incidence of second primaries than their older counterparts.
An astute caveat to these reported 5-year survival rates, as pointed out by Vargas and colleagues, is the difference between 5-year overall survival and 5-year disease-free survival. In their series of 17 female patients younger than age 40 with tongue squamous cell carcinoma (SCC) and 17 controls matched for stage, the 5-year overall survival of the study group was 65%, but the 5-year disease-free survival was only 47%.  This distinction between “alive with no evidence of disease” and “alive with disease” was not explicitly addressed in the NCDB, SEER, Scandinavian, or Washington University studies above. [16, 4, 9, 56] Young, otherwise healthy patients may be able to survive with incurable disease longer than their older counterparts, and this may confound studies based on cancer databases that do not make this distinction.
Patterns of recurrence
Several small case series have been published reporting young patients with a high rate of locoregional recurrence. [1, 57, 58, 54] These case series may be subject to publication or referral bias but do highlight a subset of young patients that may have more aggressive disease. Another explanation for the higher local recurrence rate could be a lack of appropriate initial surgical treatment. Because of their young age, these patients may not have been as aggressively treated as their older counterparts, although in the absence of details about margin assessment and extent of resection, this would be speculation.
Two small, matched control studies [32, 55] and one retrospective institutional series  also found high locoregional recurrence in younger patients compared with older patients, but none demonstrated a corresponding difference in 5-year survival. On the contrary, Von Doersten and colleagues did not find age to affect recurrence at all in a multivariate analysis on 155 patients, of whom 23 were under age 40.  Thus, whether young patients, or a subset thereof, have a higher propensity for locoregional recurrence has yet to be definitely determined.
Second primary tumors
In patients of all ages who have been treated for HNSCC, the risk of development of second primaries increases over time  and is linked to tobacco consumption.  In a matched control study, Verschuur and colleagues found a decreased incidence of second primaries among patients younger than 40 years compared with older patients (8% versus 18%) over 10 years, but this may have been confounded by the higher proportion of smokers in the older age group.  Similarly, Lacy and colleagues in a retrospective database review in which 40 of the 1030 patients were age 40 or younger, the risk of recurrence or second primary tumors was significantly lower than in the older age groups. 
Young patients may have a lower incidence of second primaries, at least in the short term. However, for those young patients who developed cancer in the absence of known risk factors, the continued presence of an unknown risk factor should be assumed. These as-of-yet unidentified risk factors may have long incubation times and patients may not present with another primary for decades after treatment of the first cancer. Only long term-studies on young patients who have been treated for HNSCC will answer these questions. In the meantime, young patients cured of their first primary deserve long-term adherence to routine follow-up and cancer surveillance, regardless of the etiology.
A study by Kam et al indicated that the suicide rate among head and neck cancer patients in the United States is more than three-fold that of the overall suicide rate. Using data from the Surveillance, Epidemiology, and End Results (SEER) program, the investigators, looking at head and neck cancer cases from between 1973 and 2011, found the age-, sex-, and race-adjusted suicide rate to be 37.9/100,000 person-years, compared with an overall US suicide rate of 11.8/100,000 person-years. The highest head and neck cancer suicide rates were for patients with laryngeal or hypopharyngeal cancer. [61, 62]
In Byers’s 1975 publication, he wrote, “the failure of physicians to vigorously investigate a persistent ulcer of the tongue in a patient less than thirty years of age resulted in considerable delay in diagnosis of many of these patients.” Enough evidence in the literature confirms that young patients can develop malignancy to warrant a high level of suspicion for cancer in any patient with worrying signs and symptoms, regardless of age.
Academic health professionals must continually update and educate our colleagues in the community regarding the identification and appropriate workup or referral for patients presenting with sings and symptoms of head and neck cancer.
Future Directions and Controversies
More than 3 decades after the first reports focusing on young patients with head and neck cancer, much remains to be learned about this subset of patients. The low incidence of head and neck squamous cell carcinoma (HNSCC) in the young patient has made development of prospective studies and analysis of single-institution data difficult. With the advent of more sophisticated and cooperative multi-institutional cancer databases, further details regarding the prognosis and treatment response of young patients may become available.
As more is learned about the association of human papillomavirus (HPV) with head and neck cancer, screening and chemopreventive strategies may have a larger role in preventing the development of HNSCC in all patients. The young patient may particularly benefit from population-based programs because they may be intercepted at an age prior to infection with HPV.
Finally, an as-of-yet-unidentified genetic defect, error in immune surveillance, or environmental exposure may be implicated in cancer development in the young patient. Future research should be directed toward identifying these risk factors.
Cancer Care Ontario guidelines
The 2009 Cancer Care Ontario guidelines for positron emission tomography (PET) scanning for head and neck cancer (updated in 2012) includes the following recommendations  :
PET scanning should be performed for the staging of head and neck squamous cell carcinoma (HNSCC) when the results of previous imaging are uncertain or treatment may be significantly modified.
When the primary cancer site is unknown, PET scanning should be performed after conventional imaging and in addition to, or prior to, diagnostic panendoscopy
PET scanning should be performed for restaging patients who are being considered for major salvage treatment, including neck dissection
In 2015, Cancer Care Ontario reaffirmed its previous guidelines for the management of stage III and IV head and neck cancer with epidermal growth factor receptor–targeted therapy. The group's recommendations are below. 
Locally advanced disease (stage III-IVB)
Recommendations include the following:
The recognized standard of care for the primary treatment of locally advanced disease is platinum-based chemoradiotherapy (CRT); 5-fluorouracil (5-FU) plus platin or monoplatin therapy appears to be most effective
In patients over age 70 years, the addition of platinum-based chemotherapy does not provide an overall survival (OS) advantage compared with radiotherapy (RT) alone
The addition of cetuximab to intensified RT may be considered as an alternative to CRT
Patients should be offered clinical trials of novel agents aimed at improving outcomes
Untreated recurrent and/or metastatic HNSCC
In recurrent and/or metastatic disease, cetuximab in combination with CRT is a superior treatment to chemotherapy alone.
Patients who were previously treated or are unsuitable for platinum-based chemotherapy
Zalutumumab may be effective against recurrent and/or metastatic HNSCC in patients who are unsuitable for cisplatin-based chemotherapy or whose disease has progressed despite the use of cisplatin-based chemotherapy.
American College of Radiology guidelines
In 2014, the American College of Radiology (ACR) released guidelines on the retreatment of HNSCC that recurs following radiation therapy. The recommendations are summarized below  :
Multidisciplinary evaluation and treatment of patients with recurrent or second primary disease is critical
For patients with technically resectable disease who are medically fit for surgery, surgical salvage is the standard of care; postoperative reirradiation with chemotherapy can be performed to improve local control and disease-free survival
Patients with metastatic disease, poor performance status, or severe toxicity from prior radiation are poor candidates for reirradiation
For patients treated with curative intent, reirradiation with chemotherapy is preferred over reirradiation alone; doses of 60 Gy or greater to the recurrent disease are recommended, but elective nodal reirradiation does not appear to be warranted; conventional fractionation or hyperfractionation with a minimum 6-hour interval is preferred
Highly conformal radiation techniques such as intensity-modulated radiation therapy (IMRT) are preferred