Re-irradiation for isolated neck recurrence in head and neck tumor: impact of rN category

Unresectable, isolated lymph node recurrence after radiotherapy is rare but a candidate for re-irradiation. However, severe toxicity is anticipated. Therefore, this study aimed to explore the efficacy and toxicity of re-irradiation in isolated lymph node recurrence of head and neck lesions. We analyzed 46 patients who received re-irradiation for lymph node recurrence without local progression. The primary tumor sites included the oral cavity in 17 patients, the hypopharynx in 12, the oropharynx in seven, the larynx in three, the nasopharynx in two, and other sites. During a median follow-up time of 10 months, the median survival time was 10.6 months, and the 1-year overall survival rate was 45.5%. The 1-year local control and progression-free survival rates were 49.8% and 39.3%, respectively. According to univariate analysis, age (≥ 65 years), the interval between treatment (≥ 12 months), rN category (rN1), and gross tumor volume (GTV < 25 cm3) were predisposing factors for better survival. In the multivariate analysis, the rN category and interval were identified as statistically significant predictors. Late toxicity grade ≥ 3 occurred in four patients (8.6%). These were all Grade 5 carotid blowout syndrome, which associated with tumor invasion of the carotid artery and/ or high doses administration for the carotid artery. Small-volume rN1 tumor that recur after a longer interval is a feasible candidate for re-irradiation. However, strict patient selection and meticulous care for the carotid are required.


Patients
We analyzed patients with isolated (without local recurrence) recurrent cervical lymph node metastases from head and neck lesions treated at seven institutions between 2002 and 2018.The inclusion criteria were as follows: (i) Re-irradiation in the same area as the previous radiotherapy, performed 30 Gy in 10 fractions or more (equivalent 2-Gy fractions = EQD2 ≥ 36 Gy, using α/β = 3 Gy), (ii) Histology confirming the pathology before the initial treatment, (iii) Recurrence after curative-intent treatment, including chemotherapy, surgery, and radiotherapy, (iv) No primary tumor progression, (v) Eastern Cooperative Oncology Group performance status scores of 0-2, (vi) Inoperable status according to the opinions of the head and neck surgeons; unresectable or medically inoperable because of severe coexisting disease or something else.
The exclusion criteria were as follows: (i) distant metastasis, (ii) palliative radiotherapy for symptomatic relief (e.g., 6-8 Gy/1 fraction), and (iii) planned boost radiotherapy and/or clinically residual cases (progression case at completion of initial radiotherapy).The study was conducted in accordance with the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board of Kyoto Prefectural University of Medicine (ERB-C-1330-3).
Three patients underwent conventional three dimensional conformal radiation therapy (3D-CRT), eight underwent intensity modulated radiotherapy (IMRT), and 35 underwent stereotactic body radiotherapy (SBRT).The gross tumor volume (GTV) was defined as visible tumor on CT/MRI images.The planning target volume (PTV) is determined by adding an adequate margin to the GTV.For example, GTV = CTV = PTV in several institutions with CyberKnife, and GTV = CTV and PTV = CTV + 2-5 mm in several institutions with LINAC.A dose of 32 Gy (median, range, 12-60 Gy) in 5 fractions (median, range, 1-30 fractions) (Table 1)   The primary endpoint was overall survival (OS).The secondary endpoints were local control (LC), progression-free survival (PFS), and toxicity grade ≥ 3. Survival data were calculated from the start of re-irradiation based on the first-event analysis for all endpoints.For these analyses, LC was defined as disease progression in the treated lymph nodes or death.PFS was defined as disease progression, relapse, or death from any cause.Toxicity was determined using the Common Terminology Criteria for Adverse Events Version 4.0.CT/MRI/ PET images were used to determine the tumor stage according to the Union for International Cancer Control (UICC) TNM Classification of Malignant Tumors, version 7.

Statistical analysis
Actuarial statistics for clinical outcomes were calculated using StatView 5.0, statistical software (SAS Institute, Cary, NC, USA) and EZR-stat 32 .A chi-square test was used to analyze the frequency.Means were compared using the Mann-Whitney U test for skewed data and the student's t-test for normally distributed data.Survival data were calculated using the Kaplan-Meier method and examined for significance using a log-rank test.Univariate and multivariate analyses of local control and survival rates were performed using Cox proportional hazards models.We examined the following factors in the univariate analysis: age, sex, location, histology, previous surgery, chemotherapy, gross tumor volume, interval between treatments, and prescribed dose.Statistically significant variables in the univariate analysis were included in the multivariate analysis.The primary tumor location (nasopharyngeal cancer or not) was omitted from this analysis because of the small number of patients with nasopharyngeal cancer (n = 2).Cutoff values were set at each variable's median value or average unless otherwise stated: 25 cm 3 for the GTV, which was set in a previous analysis 33 .Statistical significance was set at p < 0.05.

Informed consent
Informed consent was obtained from all subjects involved in the study.

Results
Table 1 summarizes the patient, tumor, and treatment characteristics.
The results of the analysis of the predisposing factors for OS are shown in Table 2. Younger age of ≤ 65 years, larger tumor size of > 25 cm 3 , advanced N category (rN2-3), and short interval between initial radiotherapy and re-irradiation ≤ 12 months were statistically significant predisposing factors for poor OS in univariate analysis (Table 2).Of these, we found that the rN category and interval were statistically significant predisposing factors for OS in multivariate analysis.
Nine patients (19%) experienced distant metastases (seven to the lungs, one to the skin, and one to the mediastinal lymph node), resulting in death.Figure 2h shows the failure patterns of the diseases.

Discussion
This study aimed to examine the re-irradiation efficacy and toxicity of isolated neck lymph node recurrences in head and neck tumors after curative radiotherapy using multi-institution data.To the best of our knowledge, this is one of the largest series of re-irradiation cohorts for rare, isolated neck lymph node recurrences.Furthermore, this is the first study to demonstrate the importance of the earlier rN category (rN1) for re-irradiation.
For re-irradiation of lymph nodes, Kawaguchi et al. reported that among the eight patients with lymph node metastases in their study, one patient with a single retropharyngeal (12.5%) had a complete response; the remaining seven patients (87.5%) all progressed 34 .Kobayashi et al. reported that the 2-year LC and OS were 81.4% and 46.3%, respectively 26 .LC was higher with a target volume ≤ 1.0 cm 3 than that with a target volume > 1.0 cm 3 (p = 0.006).Fatal bleeding was observed in one patient with a widespread tumor that invaded the carotid artery.Pollard et al. also reported good outcomes from re-irradiation for small retropharyngeal nodal metastases with a high prescribed dose without Grade ≥ 3 late toxicity 27 .Among 19 patients, the 1-year LC, locoregional control, OS, and PFS were 100%, 94%, 92%, and 92%, respectively.For entire head neck cancer re-irradiation, Diao et al. reported 44.3 months of MST with a 1-year local control rate of 78% using 45 Gy/ 9 fractions of SBRT for a median target volume of 16.9 cm 3 (small tumors) in 137 patients 34 .Our data concur with these findings and suggest that smaller lesions are associated with better outcomes.This is reflected in the traditional Union for International Cancer Control T category, which uses the diameter of oral, oropharyngeal, and hypopharyngeal cancers to classify a tumor into either T1 (-2 cm) or T2 (-4 cm).This reflects 4.1 cm 3 and 33.4 cm 3 in spherical tumor volume.In addition, we found that rN1 (ipsilateral single lymph node ≤ 3 cm) and interval were important predictors of prognosis, which concurred with the discussion for the tumor volume factor.The importance of tumor volume has been highlighted in the literature 12,[21][22][23][28][29][30]34,35 . Small-volum isolated rN1 cervical lymph node-oriented directed therapy, including re-irradiation, may play a role in some cases.We prescribe 60 Gy or more in EQD2G for small lesions (i.e., rN1) according to recent studies.
The treatment interval is also an important predisposing factor for re-irradiation in the literature 12,[21][22][23][28][29][30]34,35 . Age is someimes reported to be an important predisposing factor for radiotherapy 36 ; however, this has yet to be confirmed 35 .We speculate that younger patients showed more advanced disease characteristics than older patients (i.e., rN2-3 tumor, shorter intervals with borderline significance) (Supplemental Table S3).Thus, we believe that age does not result in a poorer prognosis in the absence of other risk factors.
We encountered four cases of CBOS, which is one of the most devastating complications of head and neck cancer treatment and mainly occurs in patients with a history of radiotherapy for tumors involving the vascular axis 31,[37][38][39][40][41][42][43][44][45][46] .Grimm et al. reported that risk factors for CBOS include a greater degree of circumferential tumor involvement around the major vessel, consecutive daily treatments, and surgical procedures before or after radiotherapy 38 .Additionally, the presence of ulceration, skin invasion, necrosis/infection, and irradiation of the lymph node area adjacent to the carotid may be risk factors for CBOS 31,[38][39][40][41][42][43][44][45][46][47][48]   www.nature.com/scientificreports/after re-irradiation is a rare [41/1,554 (2.6%)] and often fatal (75%) event 43 .We also reported that CBOS occurred in 8.4% of cases among 381 patients treated with 484 re-irradiation sessions using stereotactic radiotherapy, and 69% of these cases were fatal 31 .Embring et al. proposed a maximal accumulated EQD2 (previous radiotherapy plus re-irradiation) of 120 Gy as a dose constraint for CBOS 40 .The Turkish group reported that CBOS was not found in cases with less than 180° of carotid invasion or D0.1cc < 47.6 Gy/ 5 fractions on nonconsecutive days 41,42 .In our study, three out of four cases showed encasement of the carotid > 180°, and all cases had postoperative recurrence with the cumulative prescribed dose of EQD2 > 105 Gy.The median total summation of EQD2 for patients with CBOS was 118 Gy (range 105.36-130Gy), whereas it was 114 Gy (78.00-145.15)for patients without CBOS (p = 0.654).Our data are in line with those of Thariat et al. 44 , who suggested maintaining the maximum cumulative dose to the carotid artery below EQD2 100 Gy in patients treated with SBRT.
Recent advances have emphasized the use of immunotherapy in difficult clinical cases.In 2019, the FDA approved pembrolizumab (a PD-1 inhibitor) as a first-line treatment for patients with metastatic or unresectable recurrent head and neck squamous cell carcinoma 47 .Combining this with re-irradiation could improve effectiveness, as radiotherapy can modulate the immune system and improve immunotherapy 48 .Prior studies have demonstrated improved local control with conventional re-irradiation combined with systemic therapy [15][16][17] .
This study had several limitations.First, it included small sample size, heterogeneous tumor subsites and histologies, and a limited follow-up time.Moreover, heterogeneous tumor locations and histological findings prevented discrete conclusions and increased selection bias risk.However, our study is one of the largest on reirradiation for isolated cervical lymph node recurrence and reveals the importance of small-volume rN1 that recurred after a longer interval; therefore, the findings may be useful.

Conclusions
Although strict patient selection and meticulous care for the carotid are required (8.6% of deaths due to CBOS), small-volume rN1-isolated lymph node recurrence in head and neck lesions that recur after a longer interval may be a candidate for re-irradiation.A multidisciplinary approach for determining the optimal dose and irradiation schedule should be prospectively evaluated.

Figure 2 .
Figure 2. (a) Overall survival rate according to age.(b) Overall survival rate according to GTV.(c) Overall survival rate ac-cording to the interval.(d) Overall survival rate according to rN category.(e) Overall survival rate according to primary sites.HPC = hypopharyngeal ca.NPC = nasopharyngeal ca., OPC = oropharyngeal ca.(f) Local control rate according to rN category.(g) Progression free survival rate according to rN category.(h) Failure pattern.T = local failure, N = nodal failure, M = distant metastases.

Table
IMRT technique with TS-1 six months later for lymph node recurrence (the right upper deep jugular chain).She showed bleeding 8.5 months after reirradiation.

Table 2 .
. McDonald et al. reported that CBOS Uni-and multivariate analysis for overall survival rate using Cox proportional hazards model.Bold values indicate statistically significance.CI Confidence interval, OS Overall survival, scc Squamous cell carcinoma.

Table 3 .
Patients and treatment characteristics of four cases with carotid blow-out syndrome (CBOS).