Introduction

Head and neck squamous cell carcinoma (HNSCC) is associated with high rates of subsequent primary malignancies [1]. Researchers report an incidence of 2–6 % per year and a cumulative second primary malignancy of 36 % at 20 years among patients with HNSCC [2].

A Primary Tumor (PT) is determined as the tumor that is diagnosed first, while a Second Primary Tumor (SPT) is determined as any other malignant neoplasm that is discovered simultaneously or later [3].

Moertel et al. [4] classified the second neoplasms according to their time sequence to appear as synchronous, those that are diagnosed simultaneously or during the first 6 months since the primary tumor was diagnosed, and metachronous, those that are diagnosed after that period of time. Regarding those tumors that appear simultaneously, a PT is determined as the tumor that caused the initial symptoms.

In the 19th century, Theodor Billroth (1829–1894) was the first one to state that any patient who had overcame a neoplasm could develop another one, only if both were malignant, had different locations and their metastasis could be individualized from each other [5].

In 1932, Warren and Gates [6] stated the criteria for second neoplasms:

  • Each tumor must be histologically malignant.

  • The possibility that the tumor is a metastasis from the other must be excluded.

  • There must be no submucous connection between both tumors and others. That means that if the PT and the second neoplasm appear in close areas and both have the same histological type, and then 2 cm of healthy tissue or dysplastic mucosa must exist between them.

Differential diagnosis between metastases versus second primary malignancies remains to be one of the most challenging tasks. The identification of a metastatic lesion is approached considering clinical and genetic parameters. “Differentiating advanced tumor spread from localized tumor growth has great influence on patient prognosis and could rationally guide evolving therapeutic strategies” [7].

Distinguishing metastases from second primary tumors depends on clinical criteria (localization of the lesion, tumor stage, and disease free interval) and genetic analysis [8]. Metastatic lesions are believed to occur at advanced stage HNSCCs with the antecedent spread to cervical lymph nodes [7]. There exists as well a close relationship between the index tumor and its metastases since both loose the same alleles on chromosomal arms.

On the other hand, in 1953, Slaughter et al. proposed the field cancerisation theory, which considered that the mucosa’s repeated exposure to carcinogenic agents, such as tobacco and alcohol, would determine the outcome of new tumors in different locations and in different time [9].

Our aim is to establish the incidence of second primary tumors, the frequent associations, and the impact on the patient’s overall survival when developing a second or third neoplasm.

Methods

A retrospective review of 307 patients with head and neck malignant neoplasms was made at the Churruca Visca Medical Center from January 2002 to December 2011. Out of them, 27 patients developed a second primary tumor. The second cancer’s diagnosis was based on Warren and Gates criteria.

This research was approved by the Churruca Visca Medical Center Ethics Committee. All of the patients from this center belong to the Federal Police Force, a closed population with no more than 170.000 members. This issue became a crucial aspect when analyzing conclusions, since the number of cases reviewed with second primary tumors is not significant compared to others in the literature.

We considered every patient under treatment for a head and neck cancer during the period previously mentioned, despite their gender or age; or that had developed a second primary tumor simultaneously, synchronically or metachronically in any other location; or that had a malignant neoplasm in the same location or far from it, with a minimum follow-up of 3 years or until their death after the last neoplasm was treated.

In order to analyze the facts, the second primary tumors were grouped into synchronous and metachronous.

The following variables with possible impact on the second primary tumor’s overall survival were analyzed: age at the time the index tumor was diagnosed, gender, primary tumor’s location in the larynx, association between two squamous cell carcinoma (index and second primary tumor), radiotherapy for the primary tumor, smoking, and persistence of this habit. Smoking persistence was defined as the continuous use of addictive substances that could lead to higher morbi- mortality [10]. We considered smoking persistence as the continuation of smoking of at least 1 cigarette/day after the diagnosis of the index tumor.

It would have been interesting to consider radiotherapy as a potential causative agent for second primary tumors; unfortunately as this research was done retrospectively, incomplete data made this impossible.

Statistical analysis

For the statistical analysis, Kaplan–Meier curves were used to assess overall survival. In order to compare mortality between two groups (SPT yes or no; smoking yes or no), the Log-Rank Test was used, considering statistically significant p < 0.05. In order to analyze the variables that determine mortality (age, gender, smoking), the Cox proportional hazards regression model was used.

Results

Out of the 307 patients assessed with malignant neoplasms, 27 developed a second primary tumor representing the 8.79 % of the population. The men–women ratio was 23:4 and the mean age of appearance of the PT is 66.8 years old. As regards the time in which the second primary tumors appear, the metachronous prevailed in 96.3 % of the cases (26 patients) with a mean time of diagnosis of 67.04 months; the synchronous represented only the 3.7 % (1 patient). Only one-third primary tumor was reported.

The Primary Tumor’s location that prevailed was head and neck, and the most frequent sites that developed SPT were larynx and hypopharynx (10 patients, 37.04 %) followed by oropharynx (4 patients, 14.81 %). The locations for SPT decreasingly were lungs (29.62 %), oropharynx (22.22 %), and larynx (18.52 %). The third primary tumor was located in the bladder.

Further analysis on the primary and second primary tumor’s location showed no differences on the following associations: head and neck cancer with head and neck cancer (10 patients), and head and neck cancer with infraclavicular cancers (10 patients).

In our series, we noticed that the location of the primary tumor influenced considerably the second primary tumor’s location. When the primary tumor was situated in the larynx–hypopharynx, oropharynx, and oral cavity, the second cancer frequently took place in the lungs (Table 1).

Table 1 Associations between SPT and PT (n = 27)
Full size table

In our research, squamous cell carcinomas predominated in primary and second primary tumors. Regarding index tumors, squamous cell carcinomas represented 70.37 % of the group (52.6 % well differentiated, 31.6 % moderately differentiated, and 15.8 % poorly differentiated) and had a ratio with other histologies of 19:8. Squamous cell carcinomas in second cancers signified 81.48 % of this group, with a ratio of 22:5 with other histologies.

When comparing histological associations between PT and SPT, the association within two HNSCC prevailed (9 patients), followed by HNSCC—squamous cell carcinoma in another region (7 patients), and non-squamous cell carcinoma in another region—HNSCC (5 patients).

While estimating the different treatment modalities, radiotherapy alone or combined, was dominant in PT as well as in SPT. In primary cancers, 19 patients (70.3 %) underwent radiotherapy: 11 patients radiotherapy + surgery, 4 patients radiotherapy, 3 patients radiotherapy + chemotherapy, 1 patient radiotherapy + chemotherapy + surgery. In second malignancies, 11 patients underwent radiotherapy, most of them in association with chemotherapy (Table 2).

Table 2 Treatment Modalities
Full size table

The overall survival was 33.3 % (9 patients), with a minimum follow-up of 3 years or until their death after the last neoplasm was treated. Out of the surviving patients, 29.6 % (8 patients) were healthy and with no disease.

The cause of death in 17 patients out of 27 (62.9 %) was cancer, and in 1 patient (3.7 %) was another disease.

Out of the patients who died of cancer, in 51.9 % (14 patients) was due to the second primary tumor: either because of recurrence (25.9 %), persistence (7.4 %), or metastasis (18.5 %). Only 11.1 % (3/27 patients) died because of the primary tumor.

When comparing the overall survival among the patients who developed a second neoplasm to those who did not have a second primary tumor, there was not a statistically significant difference (Fig. 1).

Fig. 1
figure 1

Overall survival SPT vs non SPT (p < 0.1496)

Full size image

Radiotherapy on the index tumor did not modify the overall survival of patients with SPT (p = 0.8632) (Fig. 2). On the contrary, smoking and its persistence did modify the overall survival, with p = 0.0295 and p = 0.0286, respectively (Figs. 3, 4). In Fig. 3, patients who did smoke seem to have better overall survival during the first months compared to those who did not. As time goes by, the overall survival of non-smokers remains constant, while that of smokers decreases abruptly.

Fig. 2
figure 2

Impact of radiotherapy applied to the index tumor

Full size image
Fig. 3
figure 3

Smoking and survival

Full size image
Fig. 4
figure 4

Smoking persistence and survival

Full size image

Discussion

Due to advances in early diagnosis and treatment of cancer, loco-regional control was improved and recurrences were reduced; nevertheless, overall survival did not show substantial improvements. The outcome of second primary tumors and distant metastasis has increased [11, 12].

Some research has shown that patients with head and neck cancer have 3 to 7 more possibilities to develop a new primary tumor than the average population; this number could raise up to 30 times more [13].

We agree with other authors that the greatest incidence of SPT is due to an extended life rate of the population, an increased overall survival of oncological patients, an extended follow-up of treated patients with a primary tumor and an enlarged exposure to carcinogens [3].

In our series, the SPT rate was 8.79 %; this is consistent with that reported by other authors which varies from 1 to 47 %, being more frequent between 7 and 22 %.

As in the series reviewed, the metachronous tumors were more than the synchronous tumors, representing 96.3 % of our population [14].

In the only patient who had two simultaneous tumors, the one that caused the symptoms that lead to consultation was treated in the first place (advanced colon cancer) and afterwards the tumor in the oral cavity (early-stage hard palate cancer). Although these patients belong to a complex group, if their clinical general conditions allow it, an aggressive loco-regional treatment would be indicated [15].

Most authors consider that the outcome of the SPT is more frequent during the first 3 years; however, in our research, with a 12-year follow-up, the outcome risk of a new tumor was consistent with the 4–7 % annual risk of developing a SPT [14, 16].

It is considered that between 3 and 4 cases of second neoplasm appear per year; the most frequent locations are head and neck subsectors, lungs, and esophagus; less frequent locations are colon, genitourinary system, pancreas, and breast [9].

According to Licciardello et al, the primary tumors in the oral cavity seem to be related to SPT in the head and neck, while primary tumors in the larynx have to do with the outcome of SPT in the lungs. Epidemiological studies have shown that the risk of suffering lung cancer after a head and neck cancer, especially in the larynx, is from 3 to 6 times more frequent than in the general population [17].

In our series, the localization of the SPT was significantly influenced by the localization of the PT. As Alvarez Marcos et al. reported, when the PT was in the larynx, oropharynx, and oral cavity, the second neoplasm was localized more frequently in the lungs.

Consistently with the different authors considered, the dominant histology for the PT and for the SPT was the squamous cell carcinoma.

In a research based on data from the National Cancer Institute, Robinson et al. found that the second tumor in the head and neck showed a more limited extension than the index tumor in the head and neck, although the differences were not statistically significant [18].

The most frequent treatment with curative purposes on the PT and SPT was radiotherapy, alone or combined. In our series, 70.4 % of the SPT had received radiotherapy applied to the index tumor, but we could not demonstrate that this association had significantly reduced overall survival. Nevertheless, some authors state that overall survival of patients with second tumors in the head and neck was greater in those who did not receive radiotherapy as treatment modality for the primary tumor. Dolan et al. [19] reported a relationship between survival of SPT and the use of radiotherapy. In his research, the percentage of dead patients because of local recurrence of the second malignancy, when it was situated in a previously irradiated area, were 61 % compared to 30 % of those who did not receive radiotherapy.

Other authors such as Farhadieh et al. analyzed the effects of radiotherapy in laryngeal squamous cell carcinomas. In their study of 987 cases, they failed to prove any difference in overall survival between patients who received radiotherapy and those that were treated only with surgery of the primary tumor. Jones et al. also found that overall survival was not considerably different in those patients who had received radiotherapy [5, 20].

Leon et al. studied 427 squamous cell carcinomas SPT and concluded that overall survival in a 5-years follow-up of patients who had received radiotherapy in the PT and those who did not was similar (56 % radiotherapy and 60.5 % without radiotherapy).

Slaughter´s Field cancerisation theory is considered until these days the basis for the development of second tumors. Patients with squamous cell carcinomas of the head and neck develop most frequently other malignancies, synchronous, or metachronous, mainly of the upper aerodigestive tract; this might be related to the level of alcohol or tobacco intoxication [21, 22].

The repetitive exposure of the mucosa to these agents might determine the outcome of new tumors separated from the index tumor by anatomic site and time. They might produce, as well, genetic alterations because of the accumulative and persistent harm [23].

In line with what was previously mentioned and in coincidence with Kiyoto et al. and Leon et al., in our analysis we could see that smoking and its persistence have had a negative impact in the overall survival of patients with SPT [24, 25].

However, these factors alone do not explain the risk of developing a second malignancy. Many patients, including smokers and drinkers, will never develop a SPT, suggesting that genetic susceptibility may also contribute in the etiology. The identification of genetically susceptible groups would allow better perception of patients at high risk [26]. The recognition of risk markers for SPT would enable to improve initial treatment and subsequent better control. The expression of p53 protein identifies patients at higher risk of recurrences or SPT [27].

The overall survival of our series was 33.3 %; 29.6 % of them were free of disease. In similarity with López Molla et al., they reported a 33 % overall survival after 5 years of the diagnosis of the SPT, and Alvarez Marcos 30 %. We assume that the development of second malignancies necessarily worsens the prognosis compared to patients who develop a single cancer.

In a recent series of 1257 patients (595 squamous cell carcinomas of the oral cavity and 662 squamous cell carcinomas of the larynx), Lin et al. [28] reported a greater overall survival in patients with SPT of the upper aerodigestive tract. Recent investigations reinforce the fact that patients suffering from a SPT HNSCC would have a better overall survival than those with a SPT from another location (lungs or esophagus) [24, 29].

Liao et al. [30] in a research carried out in patients with carcinoma of the oral cavity perceived that SPT was diagnosed at a less advanced stage compared to index tumors. This is because of a rigorous follow-up of patients previously treated of a primary tumor in that location.

Several authors believe that early diagnosis and treatment of a second neoplasm would be critical to improve the overall survival. Despite being diagnosed at an early stage, overall survival of patients who develop a SPT is lower than that of patients with a single tumor.

Conclusions

Second primary tumors are frequent in the head and neck area, and their outcome worsens considerably the patient’s prognosis. Smoking persistence after treatment would have a negative impact on overall survival, but not the use of radiotherapy to treat the index tumor. The most common associations were larynx–lung, regarding the organ; squamous cell carcinomas, regarding the histology; and the head and neck, regarding the location.

We enforce the importance of an extended follow-up of patients successfully treated of cancer of the head and neck due to their increased risk of developing a second primary tumor.