Nasopharyngeal carcinoma (NPC) is rare in most parts of the world, but it is endemic in Southeast Asia, South China, and North Africa [1]. Radiotherapy is the primary treatment of NPC due to the close relationship with the skull base, which limits surgical accessibility and a relatively high radio-sensitivity [2]. The current treatment of NPC includes curative RT alone for early NPC and concurrent chemo-radiotherapy with or without adjuvant or neo-adjuvant chemotherapy for loco-regionally advanced disease [3].

Elderly patients aged ≥ 70 years old account for about 10% of all the patients diagnosed with NPC [4]. The treatment of elderly NPC remains challenging, as elderly patients aged ≥ 70 were often excluded from or under-presented in the clinical trials [5]. Elderly patients often present with various commodities, impaired physical and psychological functions, and compromised nutritional status [6]. There is currently no consensus regarding the RT technique, total dose, and dose fraction for elderly patients with NPC. Although intensity-modulated RT (IMRT) with curative dose is widely used in clinical practice, the use of combined chemotherapy (CT) in elderly NPC patients with loco-regional advanced disease is another controversial issue [7]. Elderly patients experience more acute and late toxicities either from the RT or a combination of RT with CT, and reports from the literature are inconsistent [8].In this study, we retrospectively reviewed our ten years of experience in a single center of treating elderly patients with NPC.

1 Materials and methods

1.1 Patients selection

This retrospective study was approved by the ethics committee of our hospital and was performed in line with the Declaration of Helsinki and Good Clinical Practice Guidelines. Patients were selected for the study from the Electronic Medical Record system (EMR) if they met all of the following criteria: pathologically confirmed nonkeratinizing NPC, aged ≥ 70 years old, and receiving at least 60 Gy IMRT. Patients with either recurrent or distant metastasis disease, synchronous metastasis, or coexisting malignancies were excluded from the study.

1.2 Evaluation

After a complete medical history and physical examination, patients were assessed by fiber-optic nasopharyngoscopy, complete blood count, blood biochemistry, and plasma Epstein-Barr virus deoxyribonucleic acid (EBV DNA) by polymerase chain reaction (PCR). Imaging study was performed, including magnetic Resonance Imaging (MRI) of the head and neck or computed tomography (CT) if the MRI was contra-indicated, chest CT, abdominal ultrasound or MRI, and bone scintigraphy.

1.3 Treatment

All patients were treated by IMRT with or without chemotherapy. The techniques of patient immobilization, target volume and critical organ delineation, and doses and dose constraints were reported else [9]. Briefly, gross tumor volumes of the primary (GTVnx) and neck nodes (GTVnd) were defined as any tumors shown in the imaging or nasopharyngoscopy. A medium-risk clinical target volume (CTV1) was defined as the GTVnx and GTVnd plus 5–8 cm margins, including the entire nasopharynx.mucosa plus a 5 mm submucosal volume, and a low-risk CTV (CTV2) was contoured as CTV1 plus 5 mm margins and any potentially involving areas for the primary tumor, which included the posterior edge of the nasal cavity and maxillary sinuses, the whole paranasopharyngeal space and retropharyngeal regions, clivus, skull base, inferior sphenoid sinus, and. pterygoid fossae. The involved neck regions were included in the CTV1, and other lymphatic drainage areas were contoured as CTV2. A total dose of 70 Gy, 60 Gy, and 54 Gy in 32–33 fractions was prescribed to the GTVnx and GTVnd, CTV1 and CTV2, respectively.

Chemotherapy was delivered at the discretion of the attending physician after a thorough discussion with the patients. The chemotherapy included concurrently single agent of cis-platin, carbo-platin, lobaplatin or nedaplatin, or a platin- based combination used adjuvantly or neo-adjuvantly.

1.4 Follow-up

The patients were evaluated weekly during the radiotherapy for tumor response and treatment toxicities. They were followed every three months in the first year, every six months during the second and third years, and yearly after three years. A physical examination, blood count and biochemistry, plasma EBV DNA detection, and nasopharyngoscopy were done at every follow-up time. An imaging study of the head and neck, chest, and liver was performed every six months or whenever indicated.

1.5 Statistics

The overall survival (OS) was measured as the time from the date of diagnosis of NPC to the date of death from any reason. The CSS was defined as the date of diagnosis to the date of death from NPC. The progression-free survival (PFS) was defined as the date of diagnosis to the date of tumor progression or death from any cause. Loco-regional recurrence-free rate (LRFR) and distant metastasis-free rate (DMFR) were defined as the date of NPC diagnosis to the time of loco-regional recurrence of distant metastasis. We used descriptive statistics to present baseline characteristics. The Kaplan–Meier method was executed to estimate the OS,CSS, PFS, LRFR, and DMFR. Cox proportional hazard regressions were performed to estimate the hazard ratio (HR). Significant variables in the univariate analyses were used in the multivariate analysis to confirm the independent prognostic factors. Variables included in the univariate comparison were age, gender, Hb level, LDH, EBV DNA, T and N classification, Staging, and treatment modality. A P value less than 0.05 was considered to be significant.

2 Results

From January 2011 to December 2020, 4351 patients with newly diagnosed NPC were admitted to our hospital, of which 221 (5.1%) were 70 years or older. After excluding 24 patients because of RT dose less than 60 Gy in 12, synchronously metastatic disease in five, receiving chemotherapy alone in six, and EGFR inhibitor alone in one. We finally included 197 patients in this study (Fig. 1).

Fig. 1
figure 1

Flow chart of patient selection

Full size image

The baseline characteristics of the selected patients are shown in Table 1. Briefly, the median age was 74 years old, 91% of the patients had stage III and IV disease, and 50% received a combination of IMRT plus CT.

Table 1 Patient and tumor characteristics
Full size table

Seventy patients died after a median follow-up of 55 months (range: 3–117 months). 31 (44%) of the deaths were directly caused by NPC progression, and 36 (51%) were due to comorbidities, malnutrition or cachexia, and treatment-related complications. The reason was unknown in 3 patients. The 5-year OS and CSS were 59.6% and 78.9%, respectively (Fig. 2A-B), and the 5-year PFS was 51.3%.

Fig. 2
figure 2

Overall survival (a) and Cancer-specific survival (b)

Full size image

There were 37 patients with tumor progression, including ten loco-regional recurrences, 24 distant metastases, and three loco-regional and distant relapses. All distant metastases and all but one loco-regional recurrence occurred four years after diagnosis. The 5-year loco-regional recurrence-free rate (LRFR) and distant metastasis-free rate (DMFR) were 91.6% and 78.9%, respectively (Fig. 3A-B).

Fig. 3
figure 3

Loco-regional recurrence-free rate (a) and distant metastasis-free rate (b)

Full size image

We performed univariate analysis by Cox proportional hazard regressions to test potential prognosticators for OS, CSS, PFS, LRFR, and DMFR. The variables included in the univariate analysis included Age (≤ 73 vs. > 73), Sex (male and female), Karnofsky performance scores (KPS, ≥ 90 vs. < 90), T category (T1-3 vs. T4), N category (N0-1 vs. N2-3), overall stage (Stage I-III vs. Stage IV), Hb level (120 vs. > 120 g/L), LDH level (≤ 250 vs. > 250 U/L), EBV DNA copies ≤ 4000 vs. > 4000 copies/ml), and treatment modality (with or without chemotherapy). As shown in Tables 2 and 3, the plasma EBV DNA was the only prognostic factor for OS, the overall staging was the only prognostic factor for CSS, and plasma EBV DNA and N category were borderline significant factors for DMFR. We did not find any prognosticator for PFS and LRFR.

Table 2 Univariate and multivariate analysis of prognosticator for OS and PFS
Full size table
Table 3 Univariate and multivariate analysis of prognosticators for LRR and DMR
Full size table

All patients developed one or more acute toxicities related to treatment. The most common acute adverse events (AE) were xerostomia, dermatitis, nausea and vomiting, mucositis, and myelosuppression. Most AE was mild, and grade 3 AE was seen in 56 (28.4%) patients. No grade 3 late toxicity in our cohort of patients.

3 Discussion

Patients ≥ 70 years old accounted for 5.1% of all the patients with NPC admitted to our hospital from January 2011 to December 2020. The proportion of elderly NPC patients in our study was lower than that reported previously. In the Hong Kong Cancer Registry report, 10.6% to 14.4% of new cases of NPC were aged ≥ 70 years [4]. In China, according to the national cancer database, 15.9% of the 11,105 newly registered NPC were aged ≥ 70 years [10]. The proportion of patients with stage III to IV disease was 91% in this study, whereas stage III to IV was approximately 75% to 80% in the literature [8]. These differences might be the result of selection bias. Elderly patients with malignancies are commonly less willing to seek medical care than their young counterparts, leading to the lower proportion of elderly patients and a more advanced stages at diagnosis.

The treatment of elderly patients was challenging. While radiotherapy remains the primary treatment for elderly NPC patients, the treatment results were suboptimal in the conventional two-dimensional era. The 5-year local control rate and OS were 63.9% and 43.4% in a cohort of patients with NPC ≥ 70 years old treated by conventional RT in a retrospective study by Zhang et al. [11]. However, IMRT improved the overall outcomes of not only the young but also the elderly patients with NPC. In the aforementioned study, Zhang et al. reported that the 5-year local control and OS were 87.3% and 67.9%. In another study by Cao et al., the 5-year loco-regional control rate, distant control rate, and OS were 92.6%, 83.7%, and 69.4% [12]. Our results show that the 5-year OS, loco-regional recurrence-free, and distant metastasis-free rates were 59.6%, 91.6%, and 78.9% and were in line with those reported in the literature.

Although loco-regional and distant tumor control in elderly patients with NPC was similar to that in young patients, the overall survival rate was only 59.6%. The sub-optimal survival might be due to the high death rate caused by non-cancer reasons, as approximately one-half of the death was caused by comorbidities, malnutrition or cachexia, and treatment-related complications. The 5-year CSS was 78.9% in our study, which was in line with the cohorts treated by IMRT, and was also similar to the young NPC patients treated in the modern era [13]. This suggested the importance of tailored treatment according to the comorbidity evaluation and geriatric assessment before treatment.

Concurrent chemo-radiotherapy with or without adjuvant or neo-adjuvant cis-platin based chemotherapy has become the standard of treatment for patients with loco-regionally advanced NPC since the publication of Intergroup 0099 study [14] and series of meta-analysis [15]. However, the use of chemotherapy in combination with radiotherapy in elderly patients with NPC is controversial. While some argued that chemotherapy should be used in elderly patients [10], many studies demonstrated no survival benefit with higher toxicities related to the chemo-radiotherapy [16]. The present study shows that incorporating chemotherapy into the treatment modality resulted in no survival and tumor control benefit. However, this should be explained with caution for its small number of cases, retrospective nature of the study, different chemotherapy used (neoadjuvant, concurrent, and adjuvant), and various platinum delivered (cis-platin, carboplatin, and nedaplatin).

In conclusion, survival after IMRT for elderly patients with NPC is suboptimal. Further study stratified by comorbidity and geriatric assessment is needed.