Tracheal cancer – treatment results, prognostic factors and incidence of other neoplasms

Primary tracheal carcinomas (TC) are rare and account for 0.01–0.4% of all malignancies with annual incidence of merely 0.2 cases per 100,000 people.^1–4 In most patients, TC are symptomatic, but often misdiagnosed as asthma, chronic pulmonary disorder or tumours of the thyroid or lungs.^3–12 Two most common histological types are squamous cell carcinoma (SCC), and adenoid cystic carcinoma (ACC) - these two types account for over 75% of primary TC.^{6, 7, 13, 14}

Knowledge on tracheal cancer is limited only to the results of retrospective reviews of institutional and national databases. Most of the studies suggest that standard methods of treatment should include extensive segmental resection with adjuvant radiotherapy (RT) or chemoradiotherapy.^{1, 3, 4–7, 13, 15} Modern techniques of tracheal surgery are associated with low perioperative risk when performed in specialized centres, and some researchers indicate that resection should be possible in more than 25–50% of patients.^{3, 14, 16, 17}

Adjuvant RT (3–6 weeks after the surgery with total dose of 60 Gy or a biologically equivalent dose of neutrons) or primary curative RT treatment in inoperable cases (doses ranging between 68 and 80 Gy) have also been proven effective in some studies.^{3, 18, 19} There is a lack of information on the late effects occurring after definitive RT for TC. The quantitative analysis of normal tissue effects in the clinic (QUANTEC) report suggests that limiting the dose to 80 Gy or lower using standard fractionation may reduce the risk of central airway stenosis. Therefore, most patients with TC are treated with lower doses (although some radioresistant tumours, such as ACC, may require higher doses to achieve local control and a small increase of the risk of treatment toxicity is acceptable).²⁰

According to Grillo et al., approximately 40% of all tracheal SCCs are reported to occur before, together with or after carcinoma of the oropharynx, larynx, or lung.¹³ In spite of the above, second cancers are not very often described among patients with TC and only one large study (270 patients) reported second cancers in 19.6% of TC patients.^{8, 13, 15, 21–23}

The aim of our study was to assess the outcome of patients treated for TC, search for potential prognostic factors, evaluate the occurrence of second malignancies, and, finally, identify those patients who potentially could have been additional surgery candidates.

We conducted a retrospective study on patients diagnosed with TC in our institution between 1997 and 2015. The Regional Cancer Database and the Hospital Database were searched for patients with tracheal neoplasms. Among 418 patients (160 from the Regional Cancer Database and 258 from the Hospital Database), we found 58 cases that met the inclusion criteria (Figure 1).

Medical records of patients were reviewed to search for second malignancies. All patients with lung or larynx cancer invading trachea were excluded from the study. Second cancer was diagnosed if one of the following criteria were met: histopathology of the second cancer was different from tracheal cancer histopathology or disease-free interval between those two cancers exceeded three years.

Additional research was performed to identify those patients, who potentially could have been candidates for surgery. Our criteria for a theoretical surgery candidate were: [1] good performance status (Eastern Cooperative Oncology Group [ECOG] 0–1), [2] no nodal or distant metastases, [3] no infiltration of surrounding organs and [4] age below 70 (due to possible surgery contraindications not associated with TC).

The overall survival (OS) was evaluated using the Kaplan-Meier method. To verify the significance of variables influencing the OS in the univariate analysis, the Cox proportional hazard model was employed. P of less than 0.05 value was considered to indicate statistical significance. Follow-up was defined as the period of time from the date of diagnosis of tracheal cancer to the date of the cut-off or death. The cut-off date for survival was December 14, 2015. Follow-up was calculated separately for the whole group and for patients who were alive at December 14, 2015. Time to progression was measured from the date of the end of the treatment to the date of local or distant progression. The study was performed according to the Helsinki Declaration and the Institutional Review Board Committee.

Epidemiological studies in Europe, USA and Japan have shown that most patients with primary malignant tracheal tumours present with an advanced local disease, have SCC tumours and are treated non-surgically.^{3, 8} Numerous authors point out that the diagnosis of TC is often delayed due to highly uncharacteristic symptoms, such as “frog in the throat” sensation reported by Howard et al.^{8, 10, 13, 18, 23} In our series, one patient had been treated with surgery and RT for thyroid carcinoma until the second histopathological examination was performed and diagnosis of TC was made. Misdiagnosis of TC as thyroid tumour was also described by other authors.^{9, 21} In their report, Licht et al. described 109 cases of TC among which 17 were diagnosed incidentally at autopsy and 21% of treated patients received asthma medication due to symptoms present at the time of diagnosis.¹³

ACC is usually diagnosed in younger patients than SCC and, in the current series, the median age of diagnosis of SCC and ACC also differ: 59 and 46, respectively.^{4, 8, 10, 13, 21, 30} ACC histology was a prognostic factor correlated with better OS in our research, which is in agreement with studies in which patients with SCC and ACC were compared.^{2, 7, 13, 15, 26, 27, 29, 31} According to these reports, five-year OS for SCC ranged from 7% to 46%, and for ACC from 10 to 89.4%. In our group, the 5-year OS for SCC and ACC were 8.4% and 77.8%, respectively.

The other factors found to correlate with better OS were: surgery or complete resection^{4, 7, 15, 27–32}, radiotherapy^{2, 4, 13, 28, 29, 31}, the status of lymphatic nodes^{7, 13, 14, 32, 33}, stage^{7, 12, 34, 35}, negative airways margins^{14, 15, 32} and the performance status.^{28, 34}

The performance status of 0 or 1 was associated with better prognosis than PS 2 in our series which is in line with previous reports. We found statistically significant influence of the smoking status on OS only in the univariate analysis, however, this difference was not statistically significant in the multivariate analysis. Other studies also did not report OS to be influenced by the smoking status.

Number of patients who had undergone surgery as the first-line treatment varied among studies from 5% to 100%.^{2, 4–8, 12–17, 25–27, 29} Licht et al., in a nationwide study of TC in Denmark, and Honings et al., in a nationwide study of TC in the Netherlands, reported that the number of patients that undergo surgery is 10% and 11.6%, respectively.^{12, 27} Honings and colleagues, in their multidisciplinary audit concerning undertreatment of TC, found 32% additional surgery candidates.⁸ They pointed out that over 50% of TC could be treated surgically if properly diagnosed and referred to a tertiary oncology centre with multidisciplinary experience in the treatment of this disease. This is confirmed in our study. Only 25% of patients were treated invasively in our series, whereas this number could rise to about 39–51%, depending on the eligibility criteria employed. Due to retrospective nature of our research, it cannot be stated that all these patients were indeed undertreated. There might have been objective factors precluding surgery which were not mentioned in the documentation available for the authors of this report. Nevertheless, this finding should be a premise to meticulous assessment of the indications and contraindications for surgery to ensure the optimal treatment of TC patients. The centralization of the care for patients with this rare airway tumour may result in selection of a greater number of patients for surgical resections and, therefore, potentially better outcomes.⁸

In many studies, surgery was followed by adjuvant radiotherapy (in 20–100% patients), and those who were not eligible for surgery received radiotherapy as the main treatment. The overall survival depended on whether surgery was performed, resection was complete or radiotherapy followed surgery.^{2, 13–15, 25–27, 29} Our findings indicate that surgical treatment or endoscopic resection of the tumour did not influence OS in the multivariate analysis. This somewhat surprising result can be explained by a bias introduced by low number of complete resections and low number of patients with any invasive procedures in our series. It is remarkable, however, that five-year OS for non-surgery and surgery/endoscopic resection group were 10% and 48%, respectively.

Radiotherapy as a sole or main treatment method was described by a number researchers.^{8, 11, 18, 19, 30–31, 33–34, 36} Only four studies present the outcome of at least 50 patients.^{8, 28, 31, 34} The median total doses usually ranged from 40 to 60 Gy₂. In our series, the doses delivered ranged from 14 to 82.6 Gy₂ and the median total dose delivered in radical radiotherapy was 66.0 Gy₂. In some reports, an additional brachytherapy boost of 6 to 30 Gy was added to escalate the dose delivered to the tumour.^{19, 28, 34, 37, 38} In our series, three patients had a brachytherapy boost and the BT total dose ranged from 12.5 to 16 Gy₂. In the studies cited above, five-year OS for patients treated only with radiotherapy ranged from 8 to 30.4%.^{19, 28, 31, 33, 34} Mornex et al. reported that the total dose over 56 (up to 70 Gy in their series) correlated with better outcomes - 12% for the higher TD compared to 5% for the lower.²⁸ Hetnał et al. reported that patients with ACC histology treated with radiotherapy had better outcomes than SCC patients, and five-year OS were 80% and 9% for ACC and SCC, respectively.³⁴ In our series, radiotherapy was a part of the treatment in 48 patients, and those patients who received RT had better overall survival in spite of large heterogeneity of the group (both radical and palliative treatments included), various fractionation schemes used, and no uniform target volume definition throughout the analysed period of time (Figure 4).

Only one of 8 patients who received brachytherapy had a haemorrhage (all others did not develop any toxicity after this treatment). Brachytherapy can be considered a part of palliative treatment like in case of bronchial infiltration in carefully selected patients. Analysis of the outcome in the current series cannot give a clear answer whether BT should be part of the treatment in all patients (for example as a boost to radical radiotherapy) because of the low number of patients in whom this treatment was employed.

There is no consensus about CTH use in TC treatment. Chemotherapy was a part of the therapy in 0% to 29% of patients in large studies.^{2, 12, 14, 25, 26, 28, 29, 31} Cisplatin was the most commonly used drug followed by cyclophosphamide and etoposide. In our series, CTH was used in 14 patients (various regimens were employed according to diagnosed histopathologic type of TC) and 13 of them (93%) received cisplatin-based therapy. CTH did not improve OS in our series which is similar to observation made by Manninen et al.³¹ However, the retrospective nature of the study and heterogeneity of the group precludes drawing unequivocal conclusions.

In the current series, TC was diagnosed as a second cancer in 10 cases, and in 1 case it was the first neoplasm. Five of the other cancers were located in the respiratory system (3 cancers of the larynx and 2 lung cancers). Vrabec found multiple primary carcinomas in 11.5% of 1518 patients presenting with upper aerodigestive cancers.³⁹ Li et al. pointed out that there is a greater likelihood of developing multiple primary carcinomas in the same system than in unrelated tissues. The head and neck region and the lung are often involved with multiple primary cancers and patients with one tumour are more likely to develop a second one.²¹ Gaissert et al. found second cancers in almost 20% of TC patients, and most of them (70%) were airway and lung cancers.¹⁵ Other authors found second cancers in 8 to 44% of patients (groups over 50 patients), mostly located in lungs or larynx.^{8, 13, 15, 21–23, 29} Histological type of the second cancer is usually the same as primary, and the SCC histology predominates.

As the number of second cancers among TC patients is relatively high, physicians engaged in the care of patients with cancer of the head and neck area should be prepared not only to treat the first malignancy, but also to be constantly vigilant for later development of subsequent primary carcinomas.³⁹

To sum up, the diagnostic workup in TC patients should be concentrated on identification of those suitable for an invasive treatment and radiotherapy as it positively influences the outcome. Respiratory system cancer survivours can be considered a risk group for the tracheal cancer. Radiotherapy is an important part of the treatment of patients with TC, whereas the role of chemotherapy requires further investigation.