The management of oropharyngeal cancer continues to evolve with new developments in biology, technology, and clinical trials. Our current series describes a cohort of over 1000 patients with stage 3 and 4 oropharynx cancer treated over the past decade. High rates of disease control were achieved overall, though analysis of variables suggests a range of prognoses dependent principally on factors at presentation, which likely reflect the HPV status of these cancers . However, it appears that treatment factors also influence outcomes, including the use of IMRT and chemotherapy in certain subgroups.
HPV status has recently been recognized as a major prognostic factor in outcomes of patients with oropharynx cancer . However, the patients in our cohort were treated prior to routine testing for HPV (and p16), and retrospective pathologic review of over 1000 samples, many of which are from outside institutions was impractical for this review, and likely to result in a large amount of missing data points.
Our cohort did consist of 423 never smokers, and it is recognized that the vast majority of patients who are never smokers present with HPV positive tumors . The 5-year local-regional control rate for these never smokers was 92% consistent with other reports on outcomes of HPV positive tumors. Without HPV status, it is more challenging to assess outcomes in HPV negative patients from our data, particularly since any history of tobacco exposure may confound the outcome of patients regardless of HPV status. Former smokers appeared to have similar outcomes to never smokers. This may be due to this cohort having a lesser intensity of smoking, but also, as noted recently by RTOG,  there is a greater percentage of HPV-positive tumors among former smokers, compared with current smokers.
Based on our past experiences, our general philosophy has been to use a risk-based approach that principally incorporates staging into our management algorithms. Historically our irradiated oropharyngeal cancer patients with T1 -2 disease had high rates of local control, [17, 26, 27] and patients who were node positive had high rates of regional control . Thus, our decision to treat patients with concurrent chemoradiation was principally reserved for those patients with bulky primary disease (T3-4). Using this approach, we observed a clear benefit for our patients with T3-4 disease treated with concurrent chemoradiation. However, despite more intensive therapy based on T-category, primary tumor size and extent remained a strong prognostic factor. Regardless of smoking status, patients with more advanced disease had a greater probability of local recurrence. More advanced T-category was also associated with a higher likelihood of distant recurrence and poorer survival.
Despite the absence of HPV stratification, our 357 patients with T1-2 disease treated with radiation without concurrent chemotherapy had 95% and 90% 5-year local regional control and overall survival rates. Current guidelines favor concurrent therapy for all stage 3 and 4 patients. The evidence for this strategy, is robust, but based on numerous studies that included patients with the most advanced disease. Particularly for patients with T1 disease, data is sparse, as few randomized trials evaluating the role of concurrent therapy included patients with low volume disease . The RTOG, for example, has excluded patients with T1 disease in their definitive trials of locally advanced head and neck cancer [2, 13, 21, 22]. Prior to the routine use of chemotherapy for head and neck cancer, many had been critical of the AJCC and UICC stage grouping ; in particular these studies have demonstrated more favorable prognoses for patients with T1 node positive disease compared with other patients staged 3-4A. Thus, particularly for T1 staged patients (who typically represent about one-third of the oropharyngeal carcinoma population) we continue to favor radiation alone.
We could not demonstrate a benefit in outcome for patients treated with induction therapy, but as described above, the rates of distant disease have decreased suggesting a benefit at least compared to historical controls. Several years ago we reported on a cohort of node positive oropharyngeal carcinoma patients treated with radiation only . These 299 patients had T1-2 disease and were treated without systemic therapy. In that report, the overall distant recurrence rate in patients with local regional control was 17%, compared with 11% for the entire current cohort here, and 7% for those staged T1-2. This observation is despite the current cohort having a higher percentage of patients staged N2b or greater, but is consistent with other recent reports of outcomes of patients treated with chemo-radiation for oropharyngeal cancer. Two principal differences between our earlier experience and the current series are that a greater percent of patients in the current series are likely HPV positive and the greater use of chemotherapy in the current experience. As many series describe similarities in distant metastases rates between HPV positive and negative patients [13, 14], greater credence is given to the second hypothesis that chemotherapy likely impacted outcomes favorably.
The retrospective nature of this study prohibits category 1 based-evidence, and retrospective studies are commonly critiqued for inherent biases. We believe the magnitude of the cohort size obviates some of the concerns regarding the inferiority of a retrospective study. In particular, all 1046 patients were treated with radiation schedules designed for curative intent. Only 8 patients (< 1%) were excluded for being treated with palliative radiation schedules, and while the patient population was extracted from a database of irradiated patients, it was extremely rare for our multidisciplinary team to treat patients without radiation.
While MDACC is a tertiary cancer center, the population presented here is likely as representative of the general population presenting with oropharynx cancer. We did not include a comorbidity index of our patients, but the population did include patients who had comorbidities that sometimes either precluded the use of cisplatin, the drug with the strongest evidence of efficacy, or chemotherapy in general, thus including patients often excluded in phase 3 trials. Comorbidities, combined with biases of patients and individuals within a large multidisciplinary team impacted ultimate treatment decisions, and help explain that while we used a risk-based approach, still some of our patients with T1 category received chemotherapy, while 10% of our patients with T4-category did not receive chemotherapy.
We also chose a relatively narrow time frame (2000 – 2007) for this study, though going further back in time would have allowed us to expand the cohort significantly. This decision was made as we wanted to establish a robust trial with adequate follow-up (median, 58 months), but still reflect on modern treatment paradigms. Despite this goal, our series had great heterogeneity of therapy which reflects on the controversies of management for this disease. The last 2 decades saw great interest in treatment intensification to improve outcomes which ironically coincided with the increase in HPV related disease which has been demonstrated to be more chemotherapy and radiation sensitive [13–15, 30]. Thus some patients were treated with chemotherapy doublets and radiation, while others received altered radiation fractionation. Favorable reports on taxane based induction regimens  led to an increased use of neoadjuvant chemotherapy, particularly for our patients with advanced nodal disease, though we frequently eliminated concurrent chemotherapy in these patients. Additionally, we often did not use chemotherapy in patients with small tumor burden despite having ‘advanced-staged’ disease. Radiation strategies changed, and we integrated more conformal radiation into our practice through the use of IMRT, and increased our use of ipsilateral radiation for selected patients with tonsillar cancer . Even with this heterogeneity of patients and treatments, we report high rates of disease control and survival.
In conclusion, we describe a cohort of over 1000 oropharyngeal cancer patients with stage 3-4B disease irradiated over an 8 year period. These patients were treated in an era in which chemotherapy was becoming well integrated into the management of advanced head and neck cancer, and IMRT developed into the routine form of radiation planning and delivery. While several trials have demonstrated HPV positivity is associated with improved prognosis, we believe T-category, a classic “biomarker” remains paramount in treatment decision-making. Thus we still advocate radiation alone often as therapy for patients with low volume disease despite nominally being categorized as having stage 3 or 4 disease. We also caution against treatment deintensification for patients with T4 disease, even if biomarkers such as HPV status suggest a more favorable prognosis. While smoking status is not a consistent surrogate for HPV status, it is clear that never smokers have excellent outcomes. Furthermore, current smokers should be strongly encouraged, counseled, and treated for cessation as an important augmentation to the principal cancer treatment.