Soft tissue metastasis in p16-positive oropharynx carcinoma: Prevalence and association with distant metastasis

Oral Oncology 51 (2015) 778–786

Contents lists available at ScienceDirect

Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology

Soft tissue metastasis in p16-positive oropharynx carcinoma: Prevalence and association with distant metastasis Parul Sinha a, James S. Lewis Jr. a,b, Dorina Kallogjeri a,c, Brian Nussenbaum a, Bruce H. Haughey a,⇑ a

Otolaryngology-Head and Neck Surgery, Washington University School of Medicine, St. Louis, MO, USA Pathology and Immunology, Washington University School of Medicine, St. Louis, MO, USA c Clinical Outcomes Research, Washington University School of Medicine, St. Louis, MO, USA b

a r t i c l e

i n f o

Article history: Received 17 February 2015 Received in revised form 11 May 2015 Accepted 12 May 2015 Available online 29 May 2015 Keywords: Oropharynx cancer Human papillomavirus p16-positive Extracapsular spread Soft tissue metastasis Adjuvant chemotherapy Adjuvant radiotherapy Distant metastasis

s u m m a r y Background: Ungraded extracapsular spread (ECS) has been found non-prognostic in p16-positive, surgically-treated oropharynx squamous cell carcinoma (OPSCC). However, soft tissue metastasis (STM), the highest ECS grade, is reported prognostic. Our study’s objective is to explore STM relative to distant metastasis (DM), the most frequent recurrence site in surgically-treated p16-positive OPSCC. Methods: Primary p16-positive OPSCC patients undergoing transoral surgery (TOS) and neck dissections were identified from a prospectively-assembled database. DM and regional recurrence (RR) rates, and DM-free survival (DMFS) were compared in pN+ patients without STM (group I) and with STM (group II). Results: Of 222 patients, 202 had pN+ disease: 147 (73%) in group I and 55 (27%) in group II. The DM rate was 6.7% (n = 15/222) overall. The DM rates were 4% (n = 6/147) vs. 16.4% (n = 9/55), RR rates were 2% (n = 3) vs. 5% (n = 3), and 5-year DMFS rates were 94.8% vs. 82.4%, in groups I and II respectively. STM was significantly associated with poorer DMFS (HR = 4.6, 95% CI: 1.65, 13.03, p = 0.004), an observation driven by its effect in the T3–T4 and not the T1–T2 subset. Amongst patients receiving adjuvant therapy, STM’s association with poorer DMFS was lost in multivariable analysis; high T-classification, however, remained significant (HR = 5.16, 95% CI: 1.43, 18.52, p = 0.012). Five-year DMFS for STM patients was 82.2% in chemoradiation (37% T3–T4) vs. 85.6% in radiation (35% T3–T4) group. Conclusions: STM was significantly associated with DM and DMFS, but only in the T3–T4, not T1–T2 subset; no significant association was seen with RR. In patients receiving adjuvant therapy, only high T-classification was associated with DMFS, not STM. Chemoradiation used as adjuvant therapy was not associated with better DMFS in STM patients for any T-classification. Ó 2015 Elsevier Ltd. All rights reserved.

Introduction The epidemiology of head and neck squamous cell carcinoma (HNSCC) in the United States has dramatically changed, now showing a high incidence of human papillomavirus (HPV)-related, p16-positive oropharynx squamous cell carcinoma (OPSCC). From cohort studies of predominantly or exclusively p16-positive, surgically-treated OPSCC, the presence of extracapsular spread (ECS) has not been associated with an adverse prognosis [1–7]. In traditional HNSCC, varying degrees of ECS are known to have an effect on disease outcomes [8–12], but even in this context, the impact of increasing degrees of ECS in p16-positive OPSCC, though categorized [1,2], has not been fully investigated or established. ⇑ Corresponding author at: Department of Otolaryngology-Head & Neck Surgery, 660 S. Euclid Avenue, Campus Box 8115, St. Louis, MO 63110, USA. Tel.: +1 (314) 362 0365; fax: +1 (314) 362 7522. E-mail address: [email protected] (B.H. Haughey). http://dx.doi.org/10.1016/j.oraloncology.2015.05.004 1368-8375/Ó 2015 Elsevier Ltd. All rights reserved.

In our two previous institutional studies [1,2], we applied an ECS grading system and designated soft tissue metastasis (STM) as the highest extent of ECS, defined as ‘‘tumor masses with no residual nodal tissue/architecture.’’ [1,2] Both studies failed to show prognostic significance for routine ECS. The first of these studies consisted of transoral surgery (TOS)- or open surgery-treated, predominantly p16-positive OPSCC patients (n = 101, 90%) [1], and the second consisted exclusively of p16-positive, TOS-treated OPSCC patients (n = 152) [2]. In the predominantly p16-positive OPSCC study, STM was prognostic for survival only at the univariate level [1]. In the exclusive p16-positive OPSCC study, STM showed an association with survival although the association was lost in patients receiving adjuvant therapy, regardless of the type (radiation alone or chemoradiation) [2]. Recurrences, although infrequent in the two studies, mainly occurred at distant sites. There is now a growing body of evidence to suggest that DM is the most common type of disease recurrence

P. Sinha et al. / Oral Oncology 51 (2015) 778–786

in surgically-treated p16-positive OPSCC patients [13–16]. The impact of STM on distant metastasis (DM) was not fully explored in the two afore-mentioned studies [1,2]. In the absence of prognostic impact from routine ECS in p16-positive OPSCC, it is necessary to explore whether the most advanced extent, or highest grade of ECS, STM, has any association with DM and thus survival. The objectives of our study were to: (a) assess the prevalence of STM in cervical nodal metastases of p16-positive OPSCC patients, with consistent STM review using clearly defined pathologic criteria, (b) explore, controlling for other relevant variables, the association between STM and DM, and to distinguish any association from lower grades of ECS, and (c) investigate any association of adjuvant therapy variables with distant metastasis-free survival (DMFS) in the presence of STM. Due to anatomic proximity, we also explored any association with regional recurrence. Materials and methods A prospectively assembled, Human Research Protection Office-approved registry for head and neck cancer was used to identify consecutively-treated OPSCC patients with TOS ± adjuvant

779

therapy at Washington University Medical School of Medicine from 1996 to 2012. The inclusion criteria were (Fig. 1): (1) Primary, biopsy-proven OPSCC treated with curative intent. (2) Extensive (>70% tumor cell) p16-positivity on immunohistochemistry (IHC) testing. (3) Minimum follow-up of 12 months or to death. (4) Availability of all histology slides from cervical nodal metastases for STM review. The slides for all patients were reviewed by our study pathologist (JSL) for ECS grading and STM detection as previously described [1,2]. Patients were excluded if all cervical nodal metastasis histology slides (as would be requisite for accurate grading) were not available. Patients with DM at presentation were excluded as were those with a prior history of HNSCC treated with surgery + adjuvant therapy or with non-surgical therapy. Patients were included if their OPSCC presented as a second primary only if the index primary was treated by TOS without any adjuvant radiation. Pertinent patient, tumor, pathology and treatment-related information were recorded from the prospective database. All data was verified and follow-up was updated.

Fig. 1. Composition of final study cohort (n = 222). TOS = transoral surgery, OPSCC = oropharyngeal squamous cell carcinoma (primary or second primary where index primary was not treated by RT), IHC = immunohistochemistry, DM = distant metastasis, ⁄incomplete therapy due to death before completion of adjuvant therapy.

780

P. Sinha et al. / Oral Oncology 51 (2015) 778–786

Tumor pathology Pathologic American Joint Committee on Cancer (AJCC) T- and N-classification were recorded from the pathology reports. p16 IHC was performed on formalin-fixed tissue specimens from primary tumor or neck nodes as previously described and was considered positive only if >70% of tumor cells showed nuclear and cytoplasmic staining, regardless of intensity [2]. The margin status, nodal metastasis size, and the presence of perineural invasion (PNI) or lymphovascular invasion (LVI) were recorded from the pathology reports. The presence of STM was determined using the nodal metastasis grading system developed by our study pathologist (JSL) as previously described [1,2]. In brief, STM was diagnosed when any of the slides from the neck dissection specimens showed a mass or masses of tumor in the soft tissue without residual organized lymph node tissue or nodal capsule (Fig. 2A and B). Size of metastasis was not a factor, such that small tumor foci in the cervical soft tissue (although uncommon) were also included as STM, particularly in the absence of larger tumor masses. A small focus of LVI is interpreted as STM only if there is microscopic presence of tumor in the neck soft tissue without an associated lymph node. Prior to 2008, the number of sections taken from grossly observed neck masses varied, with a minimum of one section for histology to several for some cases. After 2008, the pathology department’s policy was to take one section per every 2 cm of dimension for grossly observed masses. Patients with nodal metastasis (pN+) were classified into two groups: I. pN+ without STM, and II. pN+ with STM. For comparison of STM with other grades of nodal metastasis, Group I was sub-divided into: Ia. pN+ without ECS, and Ib. pN+ with ECS, without STM (Fig. 3).

Statistical analysis The primary endpoints of our exploratory study were DM rate and DMFS. DM rate was defined as the number of cases with DM over the total number in the cohort over the study period. DMFS was defined as the time from surgery to the detection of DM or death of any cause. Regional recurrence (RR) rate was the secondary end-point and was defined as recurrence in regional lymph nodes separate from any local recurrence. Heterogeneity between groups was investigated using Chi-square or Fisher’s Exact Test for categorical data and independent t test for continuous data. All statistical tests were 2-sided and were evaluated at alpha level of 0.05. Five-year Kaplan–Meier DMFS estimate was calculated and survival estimates were compared by log-rank statistic. Cox

proportional hazards (PH) multivariable models were used to adjust for covariates found to be significant predictors of DMFS in the univariable analysis. Hazard ratios (HR) with 95% confidence intervals (CI) were generated. The assumption of proportionality was tested using estimated log( log) survival plots. Data was analyzed using SPSS (IBM SPSS Statistics, Rel 20.0.0, Chicago: IBM Corporation).

Results Of 298 patients with OPSCC in the TOS registry (1996–2012), exclusions led to a final study cohort of 222 p16-positive patients, 193 males and 29 females (Fig. 1). The prognostic association of STM with survival for 152 of the current 222 patients (1996– 2010) was reported in an earlier publication [2]. For the previous report [2], 9 cases were excluded because of lack of all necessary histologic slides. However, in the interim, slides could be retrieved for 4 of these patients so they were included in the current study. Demographic, patient, pathologic and treatment details are illustrated in Table 1. Of 222 patients, 175 (79%) received adjuvant therapy – RT alone in 97 (44%) and concurrent CRT in 78 (35%). Median follow-up duration for alive patients was 60 months [minimum–maximum (min–max) = 15–189 months]. Twenty-six (11%) patients died; 16 (7%) from disease and 10 patients (4.5%) from other causes (second primary tumor = 2; medical/surgical illnesses = 7; skull base radiation necrosis 3 years post-treatment = 1).

Pathology pT-classification was T1 in 97 (44%), T2 in 70 (31%), T3 in 31 (14%) and T4 in 24 (11%) patients. pN-classification was N0 in 20 (10%), N1 in 31 (14%), N2a in 39 (17%), N2b in 93 (42%), N2c in 23 (10%), and N3 in 16 (7%). Amongst the pN+ patients (n = 202), Group I consisted of 147 (73%) patients, 107 in Group Ia and 40 in Group Ib. Group II consisted of 55 (27%) patients. The ECS grades and their distribution amongst the patient cohort are shown in Fig. 3. The demographic, tumor, and treatment characteristics for pN0 and pN+ patients by the presence or absence of STM are shown in Table 1. STM in pN+ patients was significantly associated with higher T-classification, size of nodal metastasis, LVI, PNI, and adjuvant therapy. Specifically, pN+ patients with T3/4 tumors were approximately twice more likely to have STM than for pN+ without STM patients.

Fig. 2. Hematoxylin and eosin-stained sections showing cervical nodal metastatic p16+ oropharyngeal squamous cell carcinoma with no extracapsular spread (A) and with soft tissue metastasis (B), respectively. (A = 0.5 magnification; B = 1.0 magnification).

781

P. Sinha et al. / Oral Oncology 51 (2015) 778–786 Table 1 Study characteristics within the pN0 and the pN+ patients, stratified by the absence or presence of soft tissue metastasis (STM). Variable

Category

pN0

p Value*

pN+

n = 20 (9%) n (%)

STM Absent n = 147 (66%) n (%)

STM Present n = 55 (25%) n (%)

Total

n (%)

Age

Mean ± SD Median (min–max)

58.6 ± 10.6 58 (43–81)

56 ± 9.7 56 (27.5–84)

57 ± 8.5 56 (36–75.4)

0.521

56.4 ± 9.47 56.4 (27.5–84)

Sex

Male Female

17 (85) 3 (15)

125 (85) 22 (15)

52 (94.5) 3 (5.5)

0.092

193 (87) 29 (13)

Tonsil Tongue base

10 (50) 10 (50)

72 (49) 75 (51)

26 (47) 29 (51)

Never Ever

9 (45) 11 (55)

80 (54) 67 (46)

16 (29) 39 (71)

610 p-y >10 p-y

2 (18) 11 (82)

12 (18) 55 (82)

8 (20) 31 (80)

ACE27 0-1 ACE27 2-3

0 0

130 (88) 17 (12)

47 (85.5) 8 (14.5)

T1–T2 T3–T4

13 (65) 7 (35)

119 (81) 28 (19)

35 (64) 20 (36)

N0–N2b N2c–N3

20 0

132 (90) 15 (10)

31 (56) 24 (44)

Negative Positive

19 (95) 1 (5)

135 (92) 12 (8)

51 (93) 4 (7)

Mean ± SD Median (min–max)

– –

3.49 ± 1.47 3.5 (1–8)

4.45 ± 1.77 4.3 (1–10)

Negative Positive

17 (85) 3 (5)

135 (92) 12 (8)

45 (82) 10 (18)

Negative Positive

16 (80) 4 (20)

116 (79) 31 (21)

28 (51) 27 (49)

None Radiation Chemoradiation

16 (80) 4 (20) 0

28 (19) 76 (52) 43 (29)

3 (5) 17 (31) 35 (64)

None Yes

16 (80) 4 (20)

28 (19) 119 (81)

3 (5) 52 (95)

Site

Smoking

0.875

Smoking

0.001

105 (47) 117 (53)

0.741

Comorbidity

22 (19) 95 (81) 0.567

pT-classification

197 (89) 25 (11) 0.01

pN-classification

167 (75) 55 (25) <0.001

Margin

Metastasis size (cm)à

114 (51) 108 (49)

183 (21) 39 (79) 1.000

PNI

205 (92) 17 (8) 0.001

3.79 ± 1.62 3.55 (1–10)

0.042

LVI

197 (89) 25 (11) <0.001

Adjuvant therapy

160 (72) 62 (28) – 47 (21) 97 (44) 78 (35) 0.017

47 (21) 175 (79)

Significant p-values are highlighted in bold. * Chi square or Fisher’s exact test for heterogeneity between pN+ patients with and without STM; SD = standard deviation, ACE = adult comorbidity evaluation, PNI = perineural invasion, LVI = lymphovascular invasion. à Information available for 166 patients.

Recurrence Tumor recurred in a total of 24 (11%) patients. DM was the most frequent mode of recurrence (n = 15, 63% of all patients with recurrence). DM occurred as the first site of recurrence in 13 patients and as the second site (after RR) in 2 patients. Of the 15 patients with DM, 12 died of disease while 3 are alive after metastatectomy in 2 and metastatectomy + chemotherapy in 1. RR was the first and only site of disease failure in 5 patients (2%), of which 4 occurred in the lateral neck (3 successfully salvaged) and 1 in a retropharyngeal node (died post re-chemoradiation). In another 2 patients, RR was successfully salvaged by neck dissection but was followed by DM after an interval of 8 months in both patients. Tumor recurred locally in 4 patients (2%), 2 of whom were successfully salvaged (1 with TOS + radiation, 1 with chemoradiation), and 2 of whom died of disease. Recurrences were confirmed histologically with the exception of 3 patients who refused biopsy (1 local recurrence confirmed

by physical examination and 2 DMs confirmed by serial imaging). The sites for DM were lung (n = 9), lung + chest wall (n = 1), lung + pericardium (n = 1), lung and brain (n = 1), lung and bones (n = 1), pericardium (n = 1), and liver (n = 1). Time to DM detection For all patients with DM (alone or with RR), recurrence was detected at a median follow-up of 21 months (min = 2.4, max = 56) after TOS. The median time to DM detection trended toward significance (p = 0.059) between group I (median = 28 months, min = 9.7, max = 56.4) vs. group II (median = 17 months, min = 2.4, max = 30). DM and RR by STM The DM and RR rates across the nodal groups are shown in Table 2. The DM rate in group II was 4 times that for group I, thus,

782

P. Sinha et al. / Oral Oncology 51 (2015) 778–786

Fig. 3. Classification of nodal metastases in pN+ patients (n = 202). Extracapsular spread (ECS) grade 0 = tumor within a lymph node surrounded by lymphoid tissue; grade 1 = tumor filling the subcapsular sinus with no intervening lymphoid tissue and with a thickened capsule/pseudocapsule, smooth leading edge, and no extension beyond the capsule; grade 2 = tumor extending 61 mm beyond the capsule; grade 3 = tumor extending >1 mm beyond the capsule; grade 4/soft tissue metastasis (STM) = irregular masses of tumor in the neck soft tissue with no histologic evidence of residual lymph node tissue or architecture (highest grade of extracapsular spread).

Table 2 Distribution of regional recurrences, distant metastasis, and death of disease, across all nodal groups.

pN0 Group I: pN+ without STM (grades 0–3) Group Ia: pN+ without ECS (grades 0–1) Group Ib: pN+ with ECS, without STM (grades 2– 3) Group II: pN+ with STM

Regional recurrence

Distant metastasis*

Death of disease**

Total

1 (5%) 3 (2%)

0 (0%) 6 (4%)

0 (0%) 8 (5%)

20 147

3 (3%)

4 (4%)

4 (3.7%)

107

0

2 (5%)

4 (10%)

40

3 (5%)

9 (16.4%)

8 (15%)

55

Significant p-values are highlighted in bold. STM = soft tissue metastasis, ECS = extracapsular spread. * Significant difference between group I vs. group II (,2, p = 0.003) and group Ia vs. group II (,2, p = 0.011). ** Significant difference between group I vs. group II (,2, p = 0.032).

showing a significant difference between pN+ with STM and pN+ without STM patients (p = 0.003). A significant difference in the DM rate was also observed between group II vs. group Ia (p = 0.011). The RR rate was not significantly different in the patients with STM compared to any of the other groups. Factors associated with DM and DMFS The 5-yr DMFS of 94.8% (95% CI: 91%, 99%) for patients in Group I was significantly different (p = 0.001) than for the 82.4% (95% CI: 72%, 93%) for patients in Group II. Compared to Group II patients, the 5-yr DMFS was 100% in pN0, 94.7% (95% CI: 90%, 100%) in group Ia (p = 0.003), and 94.8% (95% CI: 88%, 100%) in group Ib patients (p = 0.08) (Fig. 4). The Kaplan–Meier estimates by STM within the patient groups of T1–T2 and T3–T4 primaries are shown in Fig. 5A and B, respectively. The univariate Cox PH regression analysis identified T3–T4 vs. T1–T2 classification, STM vs. pN+ without STM, and LVI to be significantly associated with reduced DMFS (Table 3). STM was also significantly associated with DMFS when compared to

Fig. 4. Kaplan Meier estimates for distant-metastasis-free survival after TOS by nodal metastasis grading [pN+ without ECS (n = 107) and pN+ with ECS, without STM (n = 40) comprised patients in the pN+ without STM group (n = 147)]. TOS = transoral surgery; ECS = extracapsular spread; STM = soft tissue metastasis.

pN+ without ECS (Table 3). In the multivariable Cox PH regression analysis adjusting for the three above-mentioned variables found significant on the univariable Cox analysis, the significance for association was retained for T3–T4 classification (HR = 5.7, 95% CI: 1.93, 16.94, p = 0.002), and STM (HR = 3.9, 95% CI: 1.48, 10.40, p = 0.006). To investigate whether the association of STM with DMFS is affected by adjuvant therapy or not, we performed Cox PH

P. Sinha et al. / Oral Oncology 51 (2015) 778–786

783

Fig. 5. Kaplan Meier estimates for distant-metastasis-free survival after TOS by the presence or absence of STM within the (A) T1–T2 (n = 167), and (B) T3–T4 (n = 55) groups. TOS = transoral surgery; ECS = extracapsular spread; STM = soft tissue metastasis.

Table 3 Univariable Cox proportional hazard regression analysis for distant metastasis-free survival. Variable

Univariable HR (95% C.I.)

p Value

Age (continuous) Sex (female vs. male) Site (tongue base vs. tonsil) Smoker (ever vs. never) Smoking (>10 vs. 610) Comorbidity (ACE-27 2–3 vs. 0–1) pT-classification (T3–T4 vs. T1–T2) pN-classification (N2c–N3 vs. N0–N2b) Margin (positive vs. negative) STM vs. pN+ without STM* STM vs. pN+ without ECS STM vs. pN+ with ECS, without STM PNI (present vs. absent) LVI (present vs. absent) Adjuvant therapy (any vs. none) Adjuvant therapy (chemoradiation vs. radiation)

1.03 (0.98, 1.09) 0.04 (0.00, 27.58) 1.21 (0.43, 3.34) 2.46 (0.78, 7.74) 0.89 (0.19, 4.16) 1.01 (0.25, 4.88) 6.76 (2.31, 19.79) 2.88 (0.81, 10.23) 3.15 (0.89, 11.18) 4.63 (1.65, 13.03) 5.06 (1.56, 16.45) 3.58 (0.77, 16.56) 2.01 (0.57, 7.12) 5.76 (1.97, 16.86) 1.5 (0.34, 6.67) 1.03 (0.35, 3.07)

0.196 0.335 0.711 0.123 0.88 0.90 <0.001 0.101 0.075 0.004 0.007 0.103 0.28 0.001 0.593 0.96

Significant p-values are highlighted in bold. * Assessed in patients with pN+ necks; ACE = adult comorbidity evaluation, STM = soft tissue metastasis, ECS = extracapsular spread, PNI = perineural invasion, LVI = lymphovascular invasion.

regression analyses in the group of pN+ patients who received adjuvant therapy [n = 171, radiation (n = 93), chemoradiation (n = 78)]. The multivariable Cox analysis confined to this group of patients including the variables of STM, T3–T4 vs. T1–T2 classification and LVI revealed that only T3–T4 classification was significantly associated with DMFS (HR = 5.16, 95% CI: 1.43, 18.52, p = 0.012); STM lost its significance (HR = 2.81, 95% CI: 0.90, 8.76, p = 0.074). A similar observation was made with STM vs. pN+ without ECS in the model, in which only T3–T4 classification was significant (HR = 4.7, 95% CI: 1.2, 18.2, p = 0.025); STM lost its significance (HR = 3, 95% CI: 0.69, 14, p = 0.136). T-classification and STM The DM rate in the T3–T4 patients with STM (35%, n = 7/20) was 6 times that of the rate in T1–T2 patients with STM (5.7%, n = 2/35).

We also performed a nested univariable Cox PH analysis of all 154 pN+, T1–T2 patients, and found that STM vs. pN+ without STM was not significantly associated with reduced DMFS (HR = 2.47, p = 0.322, 95% CI: 0.41, 14.81). There was no significant difference in adjuvant therapy administration between the two groups, with 94% patients receiving adjuvant in the STM vs. 81% in the pN+ without STM group (p = 0.07). Amongst the pN+, T1–T2 patients receiving adjuvant therapy (n = 129), STM vs. pN+ without STM was again found to have no significant association with reduced DMFS (HR = 3.23, p = 0.241, 95% CI = 0.45, 23.04). Of note, in a univariable Cox analysis of the 55 pN+, T3–T4 patients, STM vs. pN+ without STM was significantly associated with reduced DMFS (HR = 4.14, p = 0.04, 95% CI: 1.06, 16.07). Adjuvant therapy and STM Amongst the 55 patients with STM, 35 underwent adjuvant chemoradiation of whom DM occurred in 6 (17%) while in the 17 who underwent radiation alone, DM occurred in 2 (12%); this difference in DM rate was not significant (p = 1.000). Three patients received no adjuvant therapy of whom DM occurred in one. Five-year DMFS for STM patients who received chemoradiation was 82% (95% CI: 69%, 95%) vs. 85.6% (95% CI: 67%, 100%) for those who received radiation alone (log rank, p = 0.723). T3–T4 disease was present in 37% of the chemoradiation-treated patients vs. 35% of the radiation-alone patients (non-significant difference). The Kaplan–Meier DMFS estimates for STM patients by adjuvant therapy type in overall and within the individual groups of T1–T2 and T3–T4 patients are shown in Figs. 6 and 7A and B, respectively. Discussion As defined by uniform criteria, STM was present in approximately one-quarter of our p16+ OPSCC cohort. The STM-positive group as a whole had a significantly reduced 5-year DMFS (82%) relative to the STM-negative group (95%). However, amongst T-classified subsets, the significant association of STM with DMFS was evident in the T3–T4 group but no association with DMFS

784

P. Sinha et al. / Oral Oncology 51 (2015) 778–786

Fig. 6. Kaplan–Meier estimates for distant metastasis-free survival in STM patients with adjuvant therapy after TOS (n = 52, 3 patients with STM did not receive adjuvant therapy). TOS = transoral surgery; STM = soft tissue metastasis; RT = radiation; CRT = chemoradiation.

was observed in the T1–T2 group. In patients with STM who received adjuvant therapy, regardless of the type, T3–T4 classification was the only variable significantly associated with reduced DMFS. DMFS, not the distant control estimate, was selected as the study’s primary endpoint since the latter would have included only patients with known DM as events. DM can occur as a delayed event in p16+ OPSCC [17,18], and if a patient is deceased from any non-OPSCC-related cause in the initial years after treatment, the risk of DM development would be unknown for such

individuals. Hence, DMFS was considered a more inclusive endpoint than distant control. The prevalence of STM in approximately one-quarter of our p16+ OPSCC cohort is similar to the STM prevalence found in previous non-HPV HNSCC studies which used the same criteria [19,20]. In the literature, lack of well-defined criterion obfuscates the true nature of soft tissue metastatic disease and thus, determination of STM prevalence [21]. For instance, in a study by Shah et al. in oral cavity cancer, ‘‘soft tissue involvement’’ was reported in 69 of 704 (10%) patients, separate from ECS but there were no defined criteria for ‘‘soft tissue involvement.’’ [22] Due to potential prognostic and therapeutic implications, this underscores the importance of using specific criteria, standard assessment, and standard reporting of STM in p16+ OPSCC, similar to ECS [23,24]. Despite the term ‘STM’ suggesting tumor metastasis to soft tissues without mention of a lymph node, STM, as a histologic entity, belongs on one end of the spectrum of ECS in neck nodal metastases. Most STM’s are large masses that represent tumor spread to regional lymph nodes which grow and become so extensive as to obliterate the node, leaving no residual nodal architecture for macro- or microscopic identification. This is reflected in our observations of mean nodal metastasis sizes in patients with STM vs. those with pN+ but no, or only typical ECS (Table 1). STM can be microscopic seen as irregular collections of tumor cells in the neck soft tissues adjacent to normal or to actual nodal metastases. These microscopic metastases probably are truly ‘‘soft tissue metastases,’’ in the sense that they represent lymphatic tumor emboli with subsequent spread of the tumor cells into the surrounding soft tissues. Both microscopic and macroscopic types were regarded as STM in this study, and previous reports have suggested that both patterns in neck metastases have similar biological and clinical significance [19,21]. The presence of microscopic STM has not been separated from macroscopic STM in the study literature, to date [19,21,25]. Compared to pN+ patients without STM, patients with STM had a greater proportion of ever-smokers, T3–T4 primaries, pN2c–N3, PNI and LVI. Poor performance status, poorly differentiated histology, and T4 disease were found as factors associated with STM by Violaris et al. [20]. In our study, 95% of patients with STM received adjuvant therapy compared to 81% patients without STM. Because

Fig. 7. Kaplan–Meier estimates for distant-metastasis-free survival after TOS for patients with STM by types of adjuvant therapy within the (A) T1–T2 (n = 33), and (B) T3–T4 (n = 19) groups. TOS = transoral surgery; STM = soft tissue metastasis; RT = radiation; CRT = chemoradiation.

785

P. Sinha et al. / Oral Oncology 51 (2015) 778–786

of the potential for these confounding variables to mask any potential independent effect of STM on DM and DMFS, we performed multivariable analyses. For isolated RR risk, we observed no significant association with STM. By contrast, for DM, we noted a rate of 16.4% in patients with STM compared to 4% in patients without STM. The median time to DM was found to be shorter in patients with STM (median = 17 months) compared to pN+ without STM patients (median = 28 months). Moreover, the presence of STM was associated with a HR of 4.6 (95% CI: 1.6, 13) for poorer DMFS compared to pN+ without STM patients. However, this effect of STM on DMFS was only present in patients with T3–T4 primaries (Fig. 4). Furthermore, in multivariable analysis conducted on patients who received adjuvant therapy (n = 171), STM no longer was found to associate with DMFS. The only associated factor was high T3–T4 classification. Two important corollaries ensue from this observation. First, this reinforces the increasingly confirmed concept of high T-classification, T3–T4, as the most important, albeit least controllable driver of DM, probably due to residual microscopic disease at the primary site, greater tumor exposure to vessels and/or local immune failure. Second, the burden of invasive nodal disease (of which STM represents the highest grade) appears to be significantly reduced in p16+ OPSCC patients who undergo a neck dissection, if followed by use of adjuvant radiation for residual microscopic tumor [26–28]. The apparent equivalence of adjuvant radiation alone to adjuvant radiation + systemic therapy implies that the benefit to patients from adjuvant therapy in the STM group is via the locoregional effects of the radiation on residual microscopic disease. This significant benefit from adjuvant radiation alone was seen even in the T3–T4 group with STM (Fig. 7B). The findings of the current study for STM and DMFS are similar to our preliminary findings regarding the role of STM in disease-specific, disease-free, and overall survival [1,2]. This is expected since DM was the most common mode for recurrence and disease-related death in those preliminary studies. The studies for STM in non-HPV stratified HNSCC do not investigate the implications of STM for DM [19,20]. In patients with traditional HNSCC of numerous anatomic subsites, Violaris et al. [20] noted a 5-year overall survival of 27% with STM, 33% with ECS, and 50% with no ECS [20]. They proposed that the presence of STM may be a confounding factor for poor survival, which was also seen in their T4 tumor patients. However, no multivariable analysis was performed. Jose et al. [19] found STM to have a significant adverse effect on actuarial and recurrence-free survival compared to patients with pN0 and pN+ without ECS. Similar to our study, no significant differences were observed between STM and pN+ with ECS group [19]. For the endpoint of DM rate in our p16+ OPSCC cohort, the DM rate in STM showed a 4-fold increase compared to pN+ without ECS group (p = 0.011). We did observe a 3-fold higher rate with the STM group compared to pN+ with ECS, without STM group; however, this difference was not statistically significant, probably due to the small number of events. While the cohort for our study was prospectively assembled, the effect of STM was assessed posthoc. Appropriate multivariable analyses were performed to adjust for the confounding effects of other prognostic variables, including adjuvant treatment. As a single institution study, there may be selection biases; however, all consecutive transoral surgery-treated patients with resectable OPSCC and known p16 status were included. Limitations due to low event numbers in the study population can impact the statistical power of the study. Addressing this limitation in a single-institutional setting is made difficult by the favorable biology and/or high treatment responsiveness of p16-positive disease. A greater sample size will be required for validation of our results, particularly those for DMFS in STM patients receiving different

adjuvant treatment modalities to non-inferiority of RT alone over CRT.

further

confirm

the

Conclusion Using clearly defined criteria, we identified STM in a quarter of patients who underwent surgery for p16+ OPSCC. A significant association between STM and DM, but not regional recurrence, was detected. When analyzed by T-classification however, STM was significantly associated with distant metastasis-free survival only in the T3–T4 group, not the T1–T2 group. Furthermore, in patients who received adjuvant therapy, association between STM and DMFS was lost, implying that the negative association of STM in the neck-dissected, T3–T4 group is mitigated by adjuvant therapy. In T3–T4 p16+ OPSCC disease with STM, the benefit of adjuvant radiation alone for DM appears to be equivalent to adjuvant chemoradiation, providing a basis for clinical trials. Conflict of interest statement None declared. References [1] Lewis JS, Carpenter DH, Thorstad WL, Zhang Q, Haughey BH. Extracapsular extension is a poor predictor of disease recurrence in surgically treated oropharyngeal squamous cell carcinoma. Mod Pathol 2011;24:1413–20. [2] Sinha P, Lewis JS, Piccirillo JF, Kallogjeri D, Haughey BH. Extracapsular spread and adjuvant therapy in human papillomavirus-related, p16-positive oropharyngeal carcinoma. Cancer 2012;118:3519–30. [3] Maxwell JH, Ferris RL, Gooding W, et al. Extracapsular spread in head and neck carcinoma: impact of site and human papillomavirus status. Cancer 2013;15(119):3302–8. [4] Klozar J, Koslabova E, Kratochvil V, Salakova M, Tachezy R. Nodal status is not a prognostic factor in patients with HPV-positive oral/oropharyngeal tumors. J Surg Oncol 2013;107:625–33. [5] Geiger JL, Lazim AF, Walsh FJ, et al. Adjuvant chemoradiation therapy with high-dose vs. weekly cisplatin for resected, locally-advanced HPV/p16-positive and negative head and neck squamous cell carcinoma. Oral Oncol 2014;50:311–8. [6] Rahmati R, Dogan S, Pyke O, et al. Squamous cell carcinoma of the tonsil managed by conventional surgery and postoperative radiation. Head Neck 2014. http://dx.doi.org/10.1002/hed.23679. [7] Kaczmar JM, Tan KS, Heitjan DF, et al. HPV-related oropharyngeal cancer: risk factors for treatment failure in patients managed with primary surgery (TORS). Head Neck 2014. http://dx.doi.org/10.1002/hed.23850. [8] Carter RL, Bliss JM, Soo KC, O’Brien CJ. Radical neck dissections for squamous carcinomas: pathological findings and their clinical implications with particular reference to transcapsular spread. Int J Radiat Oncol Biol Phys 1987;13:825–32. [9] Olsen KD, Caruso M, Foote RL, et al. Primary head and neck cancer. Histopathologic predictors of recurrence after neck dissection in patients with lymph node involvement. Arch Otolaryngol Head Neck Surg 1994;120:1370–4. [10] Brasilino de Carvalho M. Quantitative analysis of the extent of extracapsular invasion and its prognostic significance. A prospective study of 170 cases of carcinoma of the larynx and hypopharynx. Head Neck 1998;20:16–21. [11] Ferlito A, Rinaldo A, Devaney KO, et al. Prognostic significance of microscopic and macroscopic extracapsular spread from metastatic tumor in the cervical lymph nodes. Oral Oncol 2002;38:747–51. [12] Greenberg JS, Fowler R, Gomez J, et al. Extent of extracapsular spread: a critical prognosticator in oral tongue cancer. Cancer 2003;97:1464–70. [13] Haughey BH, Sinha P. Prognostic factors and survival unique to surgically treated p16+ oropharyngeal cancer. Laryngoscope 2012;122:S13–33. Suppl (September). [14] Haughey BH, Hinni ML, Salassa JR, et al. Transoral laser microsurgery as primary treatment for advanced-stage oropharyngeal cancer: a United States multicenter study. Head Neck 2011;33:1683–94. [15] Cohen MA, Weinstein GS, O’Malley BW, Feldman M, Quon H. Transoral robotic surgery and human papillomavirus status: oncologic results. Head Neck 2011;33:573–80. [16] Moore EJ, Olsen SM, Laborde RR, et al. Long-term functional and oncologic results of transoral robotic surgery for oropharyngeal squamous cell carcinoma. Mayo Clin Proc 2012;87:219–25. [17] Sinha P, Thorstad WT, Nussenbaum B, et al. Distant metastasis in p16-positive oropharyngeal squamous cell carcinoma: a critical analysis of patterns and outcomes. Oral Oncol 2014;50:45–51.

786

P. Sinha et al. / Oral Oncology 51 (2015) 778–786

[18] Huang SH, Perez-Ordonez B, Weinreb I, et al. Natural course of distant metastases following radiotherapy or chemoradiotherapy in HPV-related oropharyngeal cancer. Oral Oncol 2013;49:79–85. [19] Jose J, Coatesworth AP, Johnston C, MacLennan K. Cervical node metastases in squamous cell carcinoma of the upper aerodigestive tract: the significance of extracapsular spread and soft tissue deposits. Head Neck 2003;25:451–6. [20] Violaris NS, O’Neil D, Helliwell TR, Caslin AW, Roland NJ, Jones AS. Soft tissue cervical metastases of squamous carcinoma of the head and neck. Clin Otolaryngol Allied Sci 1994;19:394–9. [21] Jose J, Moor JW, Coatesworth AP, Johnston C, MacLennan K. Soft tissue deposits in neck dissections of patients with head and neck squamous cell carcinoma: prospective analysis of prevalence, survival, and its implications. Arch Otolaryngol Head Neck Surg 2004;130:157–60. [22] Shah JP, Cendon RA, Farr HW, Strong EW. Carcinoma of the oral cavity. factors affecting treatment failure at the primary site and neck. Am J Surg 1976;132: 504–7. [23] Brannan AG, Johnstone PAS, Cooper J. Extracapsular tumor extension in cervical lymph nodes: reconciling the literature and seer data. Head Neck 2011;33:525–8.

[24] Van den Brekel MWM, Lodder WL, Stel HV, Bloemena E, Leemans CR, van der Waal I. Observer variation in the histopathologic assessment of extranodal tumor spread in lymph node metastases in the neck. Head Neck 2012;34: 840–5. [25] Moor JW, Jose J, Johnston C, Coatesworth AP, MacLennan Ka. Upper aerodigestive tract squamous cell carcinoma: distribution of extracapsular spread and soft tissue deposits in the neck. Acta Otolaryngol 2004;124: 97–101. [26] Grandi C, Alloisio M, Moglia D, et al. Prognostic significance of lymphatic spread in head and neck carcinomas: therapeutic implications. Head Neck Surg 1985;8:67–73. [27] Mamelle G, Pampurik J, Luboinski B, Lancar R, Lusinchi A, Bosq J. Lymph node prognostic factors in head and neck squamous cell carcinomas. Am J Surg 1994;168:494–8. [28] Pinsolle J, Pinsolle V, Majoufre C, Duroux S, Demeaux H, Siberchicot F. Prognostic value of histologic findings in neck dissections for squamous cell carcinoma. Arch Otolaryngol Head Neck Surg 1997;123:145–8.