Abstract
Purpose
Adenoid cystic carcinoma (ACC) is a uncommon salivary malignant tumor. Our aim was to review our experience with parotid ACC, to identify clinical-pathological parameters predictive for outcome.
Materials and methods
We retrospectively reviewed 228 patients affected by parotid gland carcinomas surgically treated at our Institution. Forty-four ACC were included in this study. Multivariate analysis risk models were built to predict recurrence free probability (RFP), distant recurrence free probability (DRFP), overall survival (OS) and disease free survival (DFS).
Results
Twenty-one patients (47.7%) died from ACC and 2.3% for other causes. The 41% presented local-regional recurrence, with a regional-RFP rate of 93%, and the 34% reported distant metastases (DM). The five and ten-year OS rates were 74% and 50%, respectively.
Conclusions
Recurrences were mainly influenced by the presence of perineural invasion and nerve paralysis, whilst female gender and age < 50 were predictors for good prognosis.
1
Introduction
Adenoid cystic carcinoma (ACC) is a malignant tumor of the head and neck district, it represents approximately the 1% of all epithelial salivary carcinomas and most frequently it involves major and minor salivary glands of the oral cavity . It presents three different morphologic patterns of growth, including tubular, cribriform and solid configuration, but all of them are characterized by a locally aggressive clinical course, with infrequent regional metastases, slowly progressive and relatively indolent distant metastasis onset, with a usually fatal outcome . The estimated 5-year survival rate is approximately 36%, with a local recurrence rate that ranges from 16% to 85% in accordance with literature . Recurrence is usually a sign of incurability, and unpredictable factors such as tumor extension along perineurial sheaths and distant metastasis appearance, notably to the lungs, together with the ACC rare incidence, make this tumor treatment challenging for every physicians.
Surgery remains the mainstay of treatment for this pathology, since its response to radiotherapy and chemotherapy has always resulted unforeseeable. Complete local resection is recommended and ipsilateral cervical lymph node dissection is warranted in case of suspicious cervical lumps on clinical examination or in case of advanced local disease .
Recent reports have tried to identify clinical and pathological prognostic factors of these patients, but outcomes have generally consisted on small single-institution retrospective series or cohort of patients affected by ACC involving different anatomical head and neck sites, without focusing on a single specific region . Factors that usually emerge to influence prognosis of parotid malignancies count clinical stage, bone involvement and resection margins status, but these general variables need to be replaced by more accurate and specific characteristics that should help in assessing single patient outcome. For instance, the American Joint Committee on Cancer (AJCC) TNM staging system , according to the T-status, the N-status, and the M-status, does not add anything to single cancer patient biology, because it does not enclose more specific variables such perineurial invasion or histologic grading, tumor site, and other patients characteristics i.e.: age, gender, smoking status and/or alcohol consumption, type of surgery performed with/without adjuvant radiotherapy or chemotherapy, late distant metastasis onset . Only one multi center international study has recently published the first nomogram analysis with the aim to predict single ACC patient prognosis using different clinical and pathological parameters never established before.
On the wake of this new enthusiasm, we have retrospectively analyzed our institutional case series of 44 patients affected by ACC with the main aim to describe their epidemiological, clinical and pathological characteristics, and to correlate all of these variables to each patient outcome in terms of RFP, DRFP, OS and DFS. Secondly, a comparison between ACC and other 184 parotid carcinomas not classified as ACC has been performed to refine and distinguish specific ACC prognostic characteristics among multiple other same-site malignancies.
2
Materials and methods
2.1
Study population
Between January 1980 and December 2005, a total of 228 patients affected by parotid gland carcinomas were treated at the Otorhinolaryngology University Clinic of Florence and 44 cases out of them received the diagnosis of ACC. All the participants signed an informed consent agreement before undergoing surgery. The indications for treatment included the presence of a parotid mass with a pre-operative fine-needle aspiration cytology (FNAC) report indicative for salivary malignancy. The study was approved by the local institutional review board (IRB) committee. All patients were reviewed and retrospectively restaged in accordance with the 7th edition of American Joint Committee on Cancer AJCC and with the WHO classification .
Patient, tumor, and treatment characteristics were extracted from each patient notes. Clinical characteristics included patient gender, age (divided into three different categories: < 50 y, 50–70 y, > 70 y), cT and cN status, clinical stage, the presence or absence of clinical facial nerve paralysis and skin invasion. Tumor characteristics included the presence of perineurial invasion (PNI), type of pattern of growth (tubular, cribriform, solid) , pathological T-status (pT), pathological N-status (pT) and surgical resection margins status. Clear margins were defined as tumor-free margins ≥ 5 mm, and positive margins if ≤ 1 mm 3 . All of tissue samples were revised by a single pathologist (AF). Treatment characteristics included type of primary tumor resection (superficial parotidectomy, total parotidectomy with facial nerve preservation and radical parotidectomy), neck dissection and use of postoperative radiation and/or chemotherapy.
Local, regional and distant recurrences onset were analyzed over a 10-year follow-up period that was done by frequent protocolled outpatient controls i.e.: every 2–3 months during the first two-year post-treatment period, every 4 months during the third year of clinical control and six monthly in the fourth and fifth year after treatment. After 5 years clinical control continued yearly for life. Diagnostic imaging, despite not routinely performed, counted CAT scans of head, neck and chest.
2.2
Statistical analysis
Categorical variables were calculated in terms of frequencies and percentages for all of the 228 parotid malignancies. First, an internal analysis of 44 ACC clinical and treatment parameters was performed. Standard descriptive statistics were used to summarize data, with respect to demographic and clinical characteristics. Chi-squared test, or Fisher exact test when appropriate, were used to compare the ACC groups against the other 184 patients.
Outcome was analyzed by univariate and multivariate survival analyses, by using STATA version 12.1 (StataCorp. 2011. Stata Statistical Software: Release 12. College Station, TX: StataCorp. LP). Logistic regression was used to investigate which factors were associated with each response variables. Afterwards, multiple logistic regression analyses were performed to account for several confounding variables simultaneously. Multiple logistic regression included all variables of interest, taking into account multicollinearity and sample size.
The Kaplan–Meier method was used to estimate overall survival (OS) defined as the period from date of diagnosis to date of death or last follow-up visit, with patients censored at their last follow-up visit by patients groups and the Log-Rank test was used to compare survival curves. The same procedure was used to analyze disease-free survival (DFS), distant recurrence-free probability (DRFP), disease-specific survival (DSS), defined as the period from date of diagnosis to date of local and regional relapse, distant metastasis onset, and death from disease, respectively, were similarly censored at the last follow-up visit for all of the clinical and pathological characteristics collected. Hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated to analyze the equality of survivor functions by subgroup for the same variables. A p value < 0.05 was considered statistically significant. A two tailed p -value < 0.05 was considered significant.
2
Materials and methods
2.1
Study population
Between January 1980 and December 2005, a total of 228 patients affected by parotid gland carcinomas were treated at the Otorhinolaryngology University Clinic of Florence and 44 cases out of them received the diagnosis of ACC. All the participants signed an informed consent agreement before undergoing surgery. The indications for treatment included the presence of a parotid mass with a pre-operative fine-needle aspiration cytology (FNAC) report indicative for salivary malignancy. The study was approved by the local institutional review board (IRB) committee. All patients were reviewed and retrospectively restaged in accordance with the 7th edition of American Joint Committee on Cancer AJCC and with the WHO classification .
Patient, tumor, and treatment characteristics were extracted from each patient notes. Clinical characteristics included patient gender, age (divided into three different categories: < 50 y, 50–70 y, > 70 y), cT and cN status, clinical stage, the presence or absence of clinical facial nerve paralysis and skin invasion. Tumor characteristics included the presence of perineurial invasion (PNI), type of pattern of growth (tubular, cribriform, solid) , pathological T-status (pT), pathological N-status (pT) and surgical resection margins status. Clear margins were defined as tumor-free margins ≥ 5 mm, and positive margins if ≤ 1 mm 3 . All of tissue samples were revised by a single pathologist (AF). Treatment characteristics included type of primary tumor resection (superficial parotidectomy, total parotidectomy with facial nerve preservation and radical parotidectomy), neck dissection and use of postoperative radiation and/or chemotherapy.
Local, regional and distant recurrences onset were analyzed over a 10-year follow-up period that was done by frequent protocolled outpatient controls i.e.: every 2–3 months during the first two-year post-treatment period, every 4 months during the third year of clinical control and six monthly in the fourth and fifth year after treatment. After 5 years clinical control continued yearly for life. Diagnostic imaging, despite not routinely performed, counted CAT scans of head, neck and chest.
2.2
Statistical analysis
Categorical variables were calculated in terms of frequencies and percentages for all of the 228 parotid malignancies. First, an internal analysis of 44 ACC clinical and treatment parameters was performed. Standard descriptive statistics were used to summarize data, with respect to demographic and clinical characteristics. Chi-squared test, or Fisher exact test when appropriate, were used to compare the ACC groups against the other 184 patients.
Outcome was analyzed by univariate and multivariate survival analyses, by using STATA version 12.1 (StataCorp. 2011. Stata Statistical Software: Release 12. College Station, TX: StataCorp. LP). Logistic regression was used to investigate which factors were associated with each response variables. Afterwards, multiple logistic regression analyses were performed to account for several confounding variables simultaneously. Multiple logistic regression included all variables of interest, taking into account multicollinearity and sample size.
The Kaplan–Meier method was used to estimate overall survival (OS) defined as the period from date of diagnosis to date of death or last follow-up visit, with patients censored at their last follow-up visit by patients groups and the Log-Rank test was used to compare survival curves. The same procedure was used to analyze disease-free survival (DFS), distant recurrence-free probability (DRFP), disease-specific survival (DSS), defined as the period from date of diagnosis to date of local and regional relapse, distant metastasis onset, and death from disease, respectively, were similarly censored at the last follow-up visit for all of the clinical and pathological characteristics collected. Hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated to analyze the equality of survivor functions by subgroup for the same variables. A p value < 0.05 was considered statistically significant. A two tailed p -value < 0.05 was considered significant.
3
Results
3.1
Continuous data description
Table 1 summarizes patient and disease characteristics for the ACC groups ( n = 44) and the other histologic types of parotid malignancies rather than ACC ( n = 184).
| Variables | Level | ACC N (%) | Others N (%) | p -Value |
|---|---|---|---|---|
| Gender | Female | 23 (52) | 71 (39) | 0.098 |
| Male | 21 (48) | 113 (61) | ||
| Age | < 50 | 165 (34) | 50 (27) | 0.416 |
| 50–70 | 22 (50) | 89 (48) | ||
| > 70 | 7 (16) | 45 (25) | ||
| cT | T1 | 8 (18) | 30 (16) | 0.001 |
| T2 | 15 (34) | 79 (43) | ||
| T3 | 17 (39) | 26 (14) | ||
| T4a | 4 (9) | 49 (27) | ||
| cN | N0 | 36 (82) | 151 (82) | 0.039 |
| N1 | 7 (16) | 11 (6) | ||
| N2a | 0 (0) | 6 (3) | ||
| N2b | 0 (0) | 14 (7) | ||
| N2c | 1 (2) | 1 (1) | ||
| N3 | 0 (0) | 1 (1) | ||
| Clinical stage | 1 | 7 (16) | 28 (15) | 0.325 |
| 2 | 14 (32) | 68 (37) | ||
| 3 | 12 (27) | 29 (16) | ||
| 4 | 11 (25) | 59 (32) | ||
| Facial nerve paralysis | No | 33 (75) | 151 (82) | 0.286 |
| Yes | 11 (25) | 33 (18) | ||
| Skin invasion | No | 40 (91) | 179 (97) | 0.051 |
| Yes | 4 (9) | 5 (3) | ||
| Type of surgery | Radical parotidectomy | 11 (25) | 60 (32) | 0.723 |
| Total parotidectomy + facial nerve preservation | 32 (73) | 118 (64) | ||
| Superficial parotidectomy | 1 (2) | 6 (3) | ||
| Perineurial invasion | No | 22 (50) | 160 (87) | < 0.001 |
| Yes | 22 (50) | 24 (13) | ||
| Cribriform pattern | No | 26 (59) | ||
| Yes | 18 (41) | |||
| Tubular pattern | No | 30 (68) | ||
| Yes | 14 (32) | |||
| Solid pattern | No | 32 (73) | ||
| Yes | 12 (27) | |||
| pT | T1 | 8 (18) | 32 (17) | 0.016 |
| T2 | 16 (36) | 73 (40) | ||
| T3 | 14 (32) | 25 (14) | ||
| T4a | 6 (14) | 54 (29) | ||
| pN | N0 | 36 (82) | 136 (84) | 0.578 |
| N1 | 6 (14) | 19 (10) | ||
| N2a | 1 (2) | 7 (4) | ||
| N2b | 1 (2) | 18 (10) | ||
| N2c | 0 (0) | 2 (1) | ||
| N3 | 0 (0) | 2 (1) | ||
| Resection margins status | No | 35 (80) | 169 (92) | 0.017 |
| Yes | 9 (20) | 15 (8) | ||
| Radiotherapy | No | 18 (41) | 75 (41) | 0.986 |
| Yes | 26 (59) | 109 (59) | ||
| Chemotherapy | No | 38 (86) | 172 (93) | 0.116 |
| Yes | 6 (14) | 12 (7) | ||
| Local recurrence | No | 28 (64) | 148 (80) | 0.017 |
| Yes | 16 (36) | 36 (20) | ||
| Regional recurrence | No | 41 (93) | 162 (88) | 0.327 |
| Yes | 3 (7) | 22 (12) | ||
| Distant metastasis | No | 29 (66) | 156 (85) | 0.016 |
| Polmone | 13 (30) | 23 (12) | ||
| Ossa | 1 (2) | 4 (2) | ||
| Cerebrale | 1 (2) | 1 (1) | ||
| Death | No | 22 (50) | 113 (61) | 0.164 |
| Yes | 21 (48) | 70 (38) | ||
| Yes, other | 1 (2) | 1 (1) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
