We summarized and analyzed the medical records of 85 young patients with histologically verified HNSCC between 2000 and 2018 based on the institutional medical database of National Institute of Oncology, Budapest, Hungary. Patients diagnosed under 40 years (≤39) of age were accepted as young, according to National Comprehensive Cancer Network (NCCN) Adolescent and Young Adult (AYA) Oncology’s guideline (11). We involved patients diagnosed and primarily treated in the Multidisciplinary Head and Neck Cancer Center or Center of Radiotherapy, as well as those who were admitted for adjuvant therapy (postoperative irradiation or chemoirradiation). According to the 10th revision of International Classification of Diseases (ICD-10) lip (C00), oral cavity (C01-C06), pharynx (C09; C10; C12; C13) and larynx (C32), with histologically verified squamous cell cancer morphology codes were included in our study (12). Patients with metastases of unknown primary tumors, as well as non-HNSCC (e.g., nasopharyngeal, salivary gland tumors, sinonasal carcinomas, thyroid tumors, lymphomas) tumors were excluded. Patient data (i.e., age, gender, drug and clinical history, tobacco and alcohol consumption, tumor site, stage, grade, and nodal status at the time of the diagnosis, time of death) were extracted from the hospital information system and the National Cancer Registry. Only those who had never smoked were considered non-smokers. Drinkers were defined as alcohol consumption at or above the World Health Organization (WHO) medium risk category (>40 g alcohol for men, or >20 g for women), which constitutes approximately a bottle of beer or >2–3 dL of wine on a daily basis (13). Each tumor was staged according to the 7th edition of TNM criteria of the Union for International Cancer Control (UICC) (14) and WHO classification was used to assess histological type and grading (12). Paraffin-embedded tissue sections were also used for the confirmation of the diagnosis and immunohistochemical staining for p16^INK4a antigen with Roche CINtec® p16 (E6H4^TM) antibody for detection of this antigen (used according to the manufacturer’s recommendation, Roche, Basle, Switzerland) in the Center of Tumor Pathology Department of National Institute of Oncology. The clinicopathological data of our study population were compared with an institutional group of 140 consecutive general HNSCC patients from year 2014.

The study was conducted under the ethical permission of the Scientific and Research Ethics Committee of the Medical Research Council (approval number: BMEÜ/3719- 1 2022/EKU) in accordance with the Helsinki Declaration of 1975, as revised in 2008.

Recurrence was defined as the local or regional return of cancer after the patient was declared disease-free. Residual tumor was defined as histologically positive resection margins after surgical therapy, or tumor suspect soft tissue detected by radiological imaging (Computed tomography/CT/, Magnetic resonance imaging/MRI/, Positron emission tomography/PET/-CT) within 3 months from the end of the radio(chemo)therapy or other conservative treatment (in certain cases the residual tumor was also proved by histology). We also diagnosed locoregional residual disease if the locoregional recurrence occurred within 3 months after the first surgical treatment regardless of the surgical margin status. Overall survival (OS) was determined as months between the date of initial diagnosis and the last follow-up or death.

Chi-square test (CS) was used to compare the composition of groups. Fisher’s exact test (FE) with Bonferroni correction of the alpha values was used as a post-hoc test for multiple comparisons. Kolmogorov-Smirnov test was used to assess the normality of distributions. Student’s t-test or Mann-Whitney U-test (MWU-test) was used to compare independent variables with normal or non-normal distribution, respectively. Survival curves were depicted according to the Kaplan-Meier method (KM), significance was analyzed by log-rank statistics (LR). Univariate and multivariate analyses of prognostic factors were done using the Cox’s regression model. A p-value < 0.05 was accepted as statistically significant. KyPlot 5.0 (KyensLab Inc., Tokyo, Japan) and Statistica 14.0.1.25 (TIBCO Software Inc., Palo Alto, CO) were used for statistical analysis.

Among young patients male predominance (78.9%, male/female: 67/18) was found. The median age at the time of the diagnosis was 37 years (range 21–39) and two-third of the patients were 35–39 years old. For comparison we involved an institutional general HNSCC group (male/female: 104/36; median age: 61.5 years, range 34–88). The most common primary tumor locations were the lip and the oral cavity in contrast to the general patient population (48/85 [56.47%] vs. 27/140 [19.28%]; p < 0.001, CS). The most common subsite was the tongue in contrast to the oral cavity tumors of the general HNSCC patients (26/48, [54.1%] vs. 6/27, [22.2%], p < 0.007, CS). In the oral cavity subgroup 6 of 48 young patients (12.5%) and 3 of 27 general HNSCC patients (11.1%) presented with lip cancer, the difference was none-significant (p = 0.85, Chi-square test). Fifty-seven of 85 young patients had early (T1-T2) primary tumor status at the time of the diagnosis, which is significantly different from the control group (57/85 [67%] vs. 66/140 [47.1%], respectively, p < 0.037). Fifty-four (63.5%) had cervical lymph node involvement (N+) according to the physical examination and radiological imaging (ultrasound, CT scan, MRI). Lymph nodes were seldom involved in laryngeal cancer (N0: 83.3%, 10 of 12). None of the patients had distant metastasis (M) at the time of the diagnosis. Well-differentiated tumors are significantly more common among young patients than in the control group (55/73 [75.3%] vs. 73/119 [61.3%], respectively, p < 0.05). Detailed characteristics of all study groups are reported in Table 1.

Data on smoking and alcohol consumption could be found in the medical records of 79 young patients. Fifty-two (65.8%) of them were (current or previous) smokers, 38 (48.1%) consumed alcohol and 33 (41.7%) enjoyed both risk factors. However, smoking was significantly less prevalent in the young group than in the general HNSCC patient population (65.8% vs. 86.4% respectively, p < 0.001, CS). In contrast, the alcohol consumption was similar in the young and the general HNSCC group (48.1% vs. 55.0%, respectively, p < 0.0001 CS) (Table 1). The p16 status was tested in 65 young and 36 control patients. The rate of positivity was 20% (n = 13) and 11.1% (n = 4) respectively, no significant difference was shown between the two groups (p = 0.25, Chi-square test).

Fifty-seven of the 85 patients (67%) underwent surgical therapy, 85.4% of those with lip and oral cavity cancer, 75% with laryngeal cancer, 40% with hypopharynx, 20% with mesopharynx cancer. Forty (70.2%) of the 57 patients underwent R0 resection, among them 10 patients with close R0 resection (the free margin is < 5 mm). In 29.8% (n = 17) of patients R1 resection was performed.

Two patients (20%) with close R0 resection, and 6 patients (35.2%) with R1 resection underwent a follow-up operation. The indications for postoperative irradiation were pT3 or pT4 stage, perineural spreading, vascular invasion, extranodal extension and close or positive surgical resection margins. Altogether 42 patients (73.6%; n = 31 patients with R0; n = 11 with R1 resection) received adjuvant therapy after the surgical therapy (Table 1.).

The median follow-up time since the diagnosis of cancer was 28.5 months (range 3–228). Cumulative five-year OS for all head and neck sites was 44.2% (Figure 1A), 44.9% for lip and oral cavity (n = 47), 58.3% for oropharynx (n = 15), 0% for hypopharynx (n = 10) and 62.8% for larynx (n = 12) carcinoma. Only the subgroup with hypopharynx carcinoma differed significantly from all others (Figure 1B).

The 5-year OS was analyzed according to risk factors. Significantly better OS was observed in the alcohol abstinent group (59.6% vs. 32.7%; p = 0.0297). The 5-year OS was significantly better in patients with early, than in advanced T status (T1-2: 52.6% vs. T3-4: 26.7%; p = 0.0058) (Figure 1C) and in patients with N0 vs. N+ nodal status (65.2% vs. 32.3% respectively; p = 0.0013) (Figure 1D).

The cumulative 5-year survival for all tumor sites was significantly better in the young than in the control group (44.2% vs. 32% respectively, p = 0.005).

Univariate Cox regression model for the entire study population revealed that young age, a non-smoker or abstinent status, as well as laryngeal tumor site might predict a significantly better prognosis. Data adjustment for multivariate Cox regression model proved that from the aforementioned parameters young age, abstinence and earlier stage remained independent prognostic parameters, which were associated with more favorable outcome (Table 2). Contrarily, alcohol consumption, older age and advanced stage almost doubled the relative risk. When analyzing the young group separately, an alcohol-abstinent status, as well as advanced stage proved significant predictors of the prognosis. All the significancies and trends in the young age group were in line with those in the entire study group (Table 3).

Residual disease was reported in 42.3% (n = 36) of the 85 patients. The histological analysis of the surgical materials showed positive margins in 17 cases, and residual disease was also detected in 15 cases after the completion of other treatment modalities (e.g., definitive radio(chemo)therapy, induction chemotherapy+ radio(chemo)therapy) by radiological imaging. We also considered four further cases of locoregional recurrences to constitute residual disease, as recurrence occured within 3 months (average of 10 weeks, range: 8–12) after surgical treatment with tumor-free margin (R0) status.

Recurrent disease was observed in 24.7% (n = 21) of the 85 patients (12 local, 4 regional, 3 distant metastasis and 2 locoregional recurrence) after that the patient was defined as disease free, and occurred within an average of 24 months after the diagnosis (local: 15.14 months/range 4–48/; regional: 9 months/range 4–20/; distant: 20 months/range 14–24/). The incidence of second primary cancers (lung, oropharynx and tongue) was 4.7% (n = 4).