Head and Neck Cancers

NCCN Categories of Evidence and Consensus

Category 1: The recommendation is based on high-level evidence (e.g., randomized controlled trials) and there is uniform NCCN consensus.

Category 2A: The recommendation is based on lower-level evidence and there is uniform NCCN consensus.

Category 2B: The recommendation is based on lower-level evidence and there is nonuniform NCCN consensus (but no major disagreement).

Category 3: The recommendation is based on any level of evidence but reflects major disagreement.

All recommendations are category 2A unless otherwise noted.

The full NCCN Clinical Practice Guidelines in Oncology for Head and Neck Cancers are not printed in this issue of JNCCN, but can be accessed online at www.NCCN.org.

Clinical trials: NCCN believes that the best management for any cancer patient is in a clinical trial. Participation in clinical trials is especially encouraged.

Comorbidity

Comorbidity refers to the presence of concomitant disease (in addition to H&N cancer) that may affect the diagnosis, treatment, and prognosis for the patient.^17–19 Documentation of comorbidity is particularly important in oncology to facilitate optimal treatment selection and estimates of prognosis. Comorbidity is known to be a strong independent predictor for mortality in patients with H&N cancer,^19–26 and comorbidity also influences costs of care, use, and quality of life.^27–29 Traditional indices of comorbidity include the Charlson index¹⁸ and the Kaplan-Feinstein index and its modifications.^19,30 The Adult Comorbidity Evaluation-27 (ACE-27) is specific for H&N cancer and has excellent emerging reliability and validity.^31,32

Quality of Life

Health-related quality-of-life issues are paramount in H&N cancer. These tumors affect basic physiologic functions (e.g., the ability to chew, swallow, and breathe), the senses (taste, smell, and hearing), and uniquely human characteristics (e.g., appearance and voice). Health status describes an individual's physical, emotional, and social capabilities and limitations. Function and performance refer to how well an individual is able to perform important roles, tasks, or activities. Quality of life differs, because the central focus is on the value (determined by the patient alone) that individuals place on their health status and function.³³

A National Institutes of Health (NIH)–sponsored conference³⁴ recommended the use of patientcompleted scales to measure quality of life. For H&N cancer–specific issues, the 3 most widely accepted validated measures are: 1) the University of Washington Quality of Life scale (UW-QOL)³⁵; 2) the EORTC Quality of Life Questionnaire (EORTC-HN35)³⁶; and 3) the Functional Assessment of Cancer Therapy-Head and Neck module (FACT-HN).³⁷ The Performance Status scale for patients with H&N cancer is a clinician-rated performance scale—with a narrower focus than the previously mentioned scales—that is also widely used.³⁸

Principles of Surgery

All patients should be evaluated by an H&N surgical oncologist before treatment. In addition, multidisciplinary evaluation and treatment must be well coordinated. Evaluation, integration of therapy, assessment of resectability, primary tumor resection, margins, surgical management of cranial nerves (VII, X–XII), neck management, management of recurrences, and surveillance (including posttreatment neck evaluation) are discussed in Principles of Surgery, pages 620–624.^1,39 Resectable disease, neck dissection, postoperative management, and salvage surgery of high-risk disease are discussed in the following sections.

Resectable Versus Unresectable Disease

The term unresectable has resisted formal definition by H&N cancer specialists. The experience of the surgeon and the support available from reconstructive surgeons, physiatrists, and prosthodontists often strongly influence recommendations, especially in institutions that treat only a few patients with locally advanced H&N cancer. The NCCN Member Institutions have teams experienced in the treatment of H&N cancer and maintain the multidisciplinary infrastructure needed for reconstruction and rehabilitation. A patient's cancer is deemed unresectable if H&N surgeons at NCCN Member Institutions do not believe they can remove all gross tumor on anatomic grounds or if they are certain local control will not be achieved after surgery (even with the addition of radiotherapy to the treatment approach). Typically, these unresectable tumors densely involve the cervical vertebrae, brachial plexus, deep muscles of the neck, or carotid artery (see pages 620–624). Tumor involvement of certain sites is associated with poor prognosis (i.e., direct extension of neck disease to involve the external skin, direct extension to mediastinal structures, prevertebral fascia, or cervical vertebrae).

Unresectable tumors (i.e., tumors that cannot be removed without causing unacceptable morbidity) should be distinguished from inoperable tumors in patients whose constitutional state precludes an operation (even if the cancer could be readily resected with few sequelae). Additionally, a subgroup of patients will refuse surgical management, but these tumors should not be deemed unresectable. Although local and regional disease may be surgically treatable, patients with distant metastases are usually treated as though the primary tumor was unresectable. Thus, patient choice or a physician's expectations regarding cure and morbidity will influence or determine treatment.

Patients with resectable tumors who can also be adequately treated without surgery represent a very important group. Definitive treatment with radiotherapy alone or radiotherapy combined with chemotherapy may represent equivalent or preferable approaches to resection in these individuals. Although these patients may not undergo surgery, their tumors should not be labeled as unresectable. Their disease is usually far less extensive than disease that truly cannot be removed.

Neck Dissection

Historically, cervical lymph node (i.e., neck) dissections have been classified as “radical” or “modified radical” procedures. The less radical procedures preserved the sternocleidomastoid muscle, jugular vein, spinal accessory nerve, or selective lymph node levels. The panel prefers to classify cervical lymphadenectomy using contemporary nomenclature, thus classifying cervical lymph node dissections as either “comprehensive” or “selective.”⁴⁰ A comprehensive neck dissection is one that removes all lymph node groups that would be included in a classic radical neck dissection. Whether the sternocleidomastoid muscle, jugular vein, or spinal accessory nerve is preserved does not affect whether the dissection is classified as comprehensive. Depending on the site, comprehensive neck dissection is often recommended for N3 disease (see specific site guidelines in the algorithm and Neck Management, pages 623 and 624).

Selective neck dissections were developed based on an understanding of the common pathways for spread of H&N cancers to regional nodes (see Figure 2).^41,42 Depending on the site, selective neck dissection is often recommended for N0 disease (see specific site guidelines in the algorithm, and pages 623 and 624). To remove the nodes most commonly involved with metastases from the oral cavity, a selective neck dissection is recommended, which includes the nodes found above the omohyoid muscle (levels I–III and sometimes the superior parts of level V).^40,43 Similarly, to remove the nodes most commonly involved with metastases from the pharynx, a selective neck dissection is recommended, which includes the nodes in levels II through IV and level VI when appropriate.⁴⁰ H&N squamous cell cancer with no clinical nodal involvement rarely presents with nodal metastasis beyond the confines of an appropriate selective neck dissection (< 10% of the time).^44–46

The chief role of selective neck dissections in these NCCN Guidelines for H&N Cancers is to select patients for possible adjuvant therapy (i.e., chemo/RT or RT), although selective neck dissections may be used as treatment when neck tumor burden is low.⁴⁷ In general, patients undergoing selective neck dissection should not have clinical nodal disease; however, selective neck dissection may prevent morbidity in patients with nodal disease and may be appropriate in certain patients with N1 to N2 disease.^48–50 In the guidelines, patients with cervical node metastases who undergo operations with therapeutic intent are generally treated with comprehensive neck dissections, because often they have disease outside the bounds of selective neck dissections. Determining whether an ipsilateral or bilateral neck dissection is needed depends on tumor thickness, the extent of the tumor, and tumor site.³⁹ For example, bilateral neck dissection is often recommended for tumors at or near the midline and/or for tumor sites with bilateral drainage.

It is particularly important for patients treated nonsurgically to have careful and regular follow-up examinations by a trained H&N surgical oncologist so that any local or regional recurrence is detected early and salvage surgery (and neck dissection as indicated) can be performed. After either RT or chemoradiation, posttreatment evaluation with imaging (e.g., CT and/or MRI with contrast, PET-CT) guides the use of neck dissection (see Post Chemoradiation or RT Neck Evaluation, page 624).^51–54 If PET-CT is used for follow-up, the first scan should be performed at approximately 12 weeks after treatment to reduce the false-positive rate.^52,55

Note that a complete clinical response (i.e., clinically negative) may be defined as no visible or palpable neck disease and no radiographic findings (i.e., the absence of either focally abnormal lymph nodes or large nodes [> 1.5 cm])^51,56; a complete pathologic response requires pathologic confirmation. If a complete clinical response is seen in patients who were N0 at initial staging, all panel members recommend observation.^51,56,57 In patients who have a clinically negative neck, a negative PET-CT is 90% reliable and further imaging is optional.^58–60 Panelists also concur that any patient with residual disease or suspected progression in the neck after radiotherapy or chemoradiation should undergo a neck dissection.⁵¹

Postoperative Management of High-Risk Disease

Many factors influence survival and locoregional tumor control in patients with H&N cancer. The role of chemotherapy in the postoperative management of the patient with adverse prognostic risk factors has been clarified by 2 separate multicenter randomized trials^61,62 and a combined analysis of data from the 2 trials for patients with high-risk cancers of the oral cavity, oropharynx, or hypopharynx.⁶³

The US Intergroup trial RTOG 95-01 randomly assigned patients with 2 or more involved nodes, positive margins, or extracapsular nodal spread of tumor to undergo standard postoperative radiotherapy or the same radiotherapy plus cisplatin 100 mg/m² every 3 weeks for 3 doses.⁶² The European trial EORTC 22931 was designed using the same chemotherapy treatment and similar radiotherapy dosing, but also included as high-risk factors the presence of perineural or perivascular disease and nodal involvement at levels IV and V from an oral cavity or oropharynx cancer.⁶¹ The RTOG trial showed statistically significant improvement in locoregional control and disease-free survival but not overall survival, whereas the EORTC trial found significant improvement in survival and the other outcome parameters. A schedule using cisplatin at 50 mg intravenously weekly has also shown improved survival in this setting.⁶⁴

View Full Size

Level designation for cervical lymphatics in the right neck.

Reprinted with permission, from CMP Healthcare Media. Source: Cancer Management: A Multidisciplinary Approach, 9th ed. Pazdur R, Coia L, Hoskins W, et al., eds. Chapter 4. Copyright 2005, All rights reserved.

Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 9, 6; 10.6004/jnccn.2011.0053

To better define risk, a combined analysis of prognostic factors and outcomes from the 2 trials was performed. This analysis showed that patients in both trials with extracapsular nodal spread of tumor and/or positive resection margins benefited from the addition of cisplatin to postoperative radiotherapy. For those with multiple involved regional nodes without extracapsular spread, no survival advantage was seen.⁶³ The panel noted that the combined analysis was considered exploratory by the authors because it was not part of the initial protocol design.⁶³ These publications form the basis of the NCCN recommendations.

In NCCN Member Institutions, patients with extracapsular nodal spread and/or positive surgical margins undergo adjuvant chemoradiotherapy after resection.^64–70 The presence of other adverse risk factors—multiple positive nodes (without extracapsular nodal spread), vascular/lymphatic/perineural invasion, pT3 or pT4 primary, and oral cavity or oropharynx primary cancers with positive level IV or V nodes—are established indications for postoperative radiotherapy. Because patients with these other adverse features were also included in the EORTC 22931 trial that showed a survival advantage for patients receiving cisplatin concurrent with postoperative radiotherapy compared with radiotherapy alone, the panel added “consider chemoradiation” for these features.⁶¹

Salvage Surgery

Patients with advanced carcinoma (any T, N2–3) who undergo nonsurgical treatment, such as concurrent chemotherapy and radiotherapy, need very close follow-up to evaluate for both local recurrence and to assess for ipsilateral or contralateral neck recurrence (see follow-up recommendations in the algorithm). Patients who do not have a complete clinical response to chemotherapy/radiotherapy require salvage surgery plus neck dissection as indicated. However, all panelists emphasized that it may be difficult to detect local or regional recurrence because of radiation-related tissue changes, and this may result in a delayed diagnosis of persistent or recurrent disease.

The panelists also emphasized the increased risk of complications when salvage surgery is attempted. Some of these patients may require microvascular free flap reconstruction to cover the defects at the primary site. The patients undergoing neck dissection may develop complications related to delayed wound healing, skin necrosis, or carotid exposure. Laryngectomy may be required to obtain clear surgical margins or prevent aspiration (e.g., in patients with advanced oropharyngeal cancer). The patients requiring salvage laryngectomy may have a high incidence of pharyngocutaneous fistula and may require either a free flap reconstruction of the laryngopharyngeal defect, or a myocutaneous flap to buttress the suture line if the pharynx can be closed primarily.

Radiation Doses

Selection of radiation total dose depends on the primary tumor and neck node size, fractionation, and clinical circumstances, including whether to use concurrent chemotherapy (see page 625 and the individual Principles of Radiation Therapy guidelines for each primary site). In general, the primary tumor and gross adenopathy require a total of 66 to 74 Gy (2.0 Gy/fraction) and up to 81.6 Gy (1.2 Gy/fraction) in hyperfractionation. External radiation doses exceeding 75 Gy using conventional fractionation (2.0 Gy/fraction) may lead to unacceptable rates of normal tissue injury.

In contrast, elective irradiation to low- and intermediate-risk nodal stations in the neck requires 44 to 64 Gy, depending on the estimated level of tumor burden and fraction size. Postoperative irradiation is recommended based on stage, histology, and surgical–pathologic findings. In general, postoperative radiotherapy is recommended for selected risk factors, including advanced T stage, depth of invasion, multiple positive nodes (without extracapsular nodal spread), or perineural/lymphatic/vascular invasion. Higher doses of radiation alone (60–66 Gy) or with chemotherapy are recommended for the high-risk features of extracapsular disease or positive margins. The preferred interval between resection and commencement of postoperative radiotherapy is 6 weeks or less.

Fractionation in Radiotherapy Alone

No single fractionation schedule has proven to be best for all tumors. Data strongly indicate squamous cancers of the H&N can grow rapidly, and may compensate for radiotherapy-induced cell loss through the mechanism of accelerated repopulation.^71–73 Especially in radiotherapy-alone settings, schedules delivering at least 1000 cGy/wk are recommended.^74–78

Two large, randomized clinical trials from Europe have reported improved locoregional control using altered fractionation. The EORTC protocol 22791 compared hyperfractionation (1.15 Gy twice daily, or 80.5 Gy over 7 weeks) with conventional fractionation (2 Gy once daily, or 70 Gy over 7 weeks) in the treatment of T2, T3, N0–1 oropharyngeal carcinoma excluding base of tongue primaries. At 5 years, a statistically significant increase was seen in local control in the hyperfractionation arm (38% vs. 56%; P = .01) and no increase in late complications.⁷⁹ A long-term follow-up analysis has also shown a small survival advantage for hyperfractionation (P = .05).⁸⁰ Another EORTC protocol (22851) compared accelerated fractionation (1.6 Gy 3 times daily, or 72 Gy over 5 weeks) with conventional fractionation (1.8–2.0 Gy once daily, or 70 Gy over 7–8 weeks) in various intermediate to advanced H&N cancers (excluding cancers of the hypopharynx). Patients in the accelerated fractionation arm had significantly better locoregional control at 5 years (P = .02). Disease-specific survival showed a trend favoring the accelerated fractionation arm (P = .06). Acute and late toxicity were increased with acceleration, however, raising questions about the net advantages of accelerated fractionation.⁸¹

The RTOG reported the initial 2-year results and subsequent mature results (after a median follow-up of 8.5 years) of a 4-armed phase III randomized clinical trial (protocol 90-03) comparing hyperfractionation and 2 variants of accelerated fractionation against standard fractionation.^82,83 After 2 years of follow-up, both accelerated fractionation with a concomitant boost (AFX-C) and hyperfractionation were associated with improved locoregional control and disease-free survival compared with standard fractionation. However, acute toxicity was increased. No significant difference was seen among the various treatment groups in the frequency of grade 3 or worse late effects reported at 6 to 24 months after treatment initiation. Long-term follow-up confirmed a statistically significant improvement in locoregional control with either AFX-C or hyperfractionation compared with standard fractionation. However, neither disease-free survival nor overall survival was significantly improved.

A meta-analysis of updated individual patient data from 15 randomized trials analyzing the effect of hyperfractionated or accelerated radiotherapy on survival of patients with H&N cancer has been published.⁸⁴ Standard fractionation constituted the control arm in all of the trials in this meta-analysis. An absolute survival benefit of 3.4% at 5 years (hazard ratio [HR], 0.92; 95% CI, 0.86–0.97; P = .003) was reported. This benefit, however, was limited to patients younger than 60 years.⁸⁴ Consensus regarding altered fractionation schedules with concomitant boost or hyperfractionation for stage III or IV oral cavity and oropharyngeal cancers has not yet emerged among NCCN Member Institutions.^83–86

Fractionation in Concurrent Chemoradiation

No consensus exists regarding the optimal radiation dose-fractionation scheme when administered with concurrent chemotherapy. Most published studies have used conventional fractionation at 2.0 Gy per fraction to 70 Gy or more in 7 weeks with single-agent cisplatin given every 3 weeks at 100 mg/m². Other fraction sizes (e.g., 1.8 Gy, conventional), other dosing schedules of cisplatin, other single agents, multiagent chemotherapy, and altered fractionation with chemotherapy have been evaluated alone or in combination. Numerous trials have shown that modified fractionation and concurrent chemotherapy are more efficacious than modified fractionation alone.^86–89 However, whether modified fractionation with concurrent chemotherapy is superior to standard fractionation and concurrent chemotherapy is currently unknown. RTOG 0129 is assessing accelerated fractionation versus standard fractionation with concurrent cisplatin. Preliminary results suggest that accelerated fractionation does not improve survival over standard fractionation.⁹⁰

Concurrent chemoradiation increases acute toxicity compared with radiation alone, although an increase in late toxicity beyond that caused by radiotherapy alone is less clear.^91–93 Altered fractionation and/or multiagent chemotherapy may further increase the toxicity burden.⁹⁴ For any chemotherapeutic approach, close attention should be paid to published reports for the specific chemotherapy agent, dose, and schedule of administration. Chemoradiation should be performed by an experienced team and should include substantial supportive care.

Radiation Techniques and IMRT

The intensity of the radiation beam can be modulated to decrease doses to normal structures without compromising the doses to the cancer targets (http://www.icru.org/index.php?option=com_content&task=view&id=171).^95,96 IMRT is an advanced form of conformal radiotherapy permitting more precise targeting of cancer while reducing dose to normal tissues.^97–100 Xerostomia is a common long-term side effect of radiotherapy, which can be reduced with IMRT, drug therapy (e.g., pilocarpine, cevimeline), and other novel approaches (e.g., acupuncture).^101–105

IMRT dose painting refers to the method of assigning different dose levels to different structures within the same treatment fraction (e.g., 2.0 to gross tumor, 1.7 to microscopic tumor, and < 1.0 Gy to parotid gland), resulting in different total doses to different targets (e.g., 70 Gy, 56 Gy, < 26 Gy).¹⁰⁶ Although dose painting has been used to simplify radiation planning, hot spots associated with higher toxicity can occur.^106,107 Alternatively, separate dose plans for the low versus higher dose targets can be delivered sequentially (“reduce target size and boost”) or on the same day as separate fractions in twice-daily schemes (see Radiation Techniques, page 625).^99,108

IMRT is now widely used in H&N cancer and is the predominant technique used at NCCN Member Institutions. It is useful in reducing long-term toxicity in oropharyngeal and nasopharyngeal cancers through reducing the dose to one or more major salivary glands, temporal lobes, mandible, auditory structures (including cochlea), and optic structures.^{101,102,109–115} However, overall survival is similar among patients treated with IMRT and those receiving conventional RT.^109,116,117 In-field recurrences, low-grade mucositis in areas away from the cancer targets, and posterior neck hair loss can occur with IMRT.^118,119 The application of IMRT to other sites (e.g., oral cavity, hypopharynx) is evolving and may be used at the discretion of treating physicians.^120,121

Numerous phase II studies show a decrease in late toxicity (xerostomia) without compromising tumor control for nasopharyngeal and other sites. More recently, 3 randomized trials have supported the clinical benefits of IMRT in H&N cancer with regard to the reduction in xerostomia. Pow et al.¹⁰¹ evaluated treatment of earlystage nasopharyngeal carcinoma with conventional radiotherapy techniques versus IMRT. The results showed a statistical improvement in salivary flow and patientreported quality-of-life parameters.¹⁰¹ In the study by Kam et al.,¹⁰² patients with nasopharyngeal carcinoma were randomized to either IMRT or conventional 2-dimensional radiotherapy. At 1 year after treatment, patients in the IMRT arm had significantly lower rates of clinician-rated severe xerostomia than patients in the 2-dimensional radiotherapy arm (39.3% vs. 82.1%; P = .001). Salivary flow rates were also higher with IMRT. The mean parotid dose was 32 Gy in the IMRT group and 62 Gy in the conventional group. Although a trend for improvement in patient-reported dry mouth was observed after IMRT, recovery was incomplete and no significant difference was seen in patient-reported outcomes between the arms. The authors concluded that other salivary glands may also be important and merit protection.

Recent data from the PARSPORT phase III randomized trial indicate that IMRT decreases xerostomia compared with conventional radiotherapy in patients with non-nasopharyngeal carcinoma.^122,123 In this trial, patients with T1–T4, N0–N3, M0 disease were treated with either conventional radiotherapy (i.e., parallel opposed technique) or IMRT to a total dose of 60 or 65 Gy in 30 fractions; 80 patients with oropharyngeal and 14 with hypopharyngeal tumors were included. Grade 2 or worse (LENT-SOMA scale) xerostomia 1 year after treatment was seen in 74% of patients receiving conventional radiotherapy versus 38% of patients in the IMRT group (P = .003). No differences were seen in the rates of locoregional control or survival.

Follow-Up After Radiotherapy

For patients whose cancer has been treated with radiotherapy, the recommended follow-up includes an assessment of thyroid function (i.e., the thyroid stimulating hormone [TSH] level should be determined every 6–12 months). Increased TSH levels have been detected in 20% to 25% of patients who received neck irradiation.¹²⁴

Brachytherapy

Brachytherapy is used less often in recent years because of improved local control obtained with concurrent chemoradiation.

Workup and Staging

Imaging studies to evaluate mandibular involvement and a careful dental evaluation (including Panorex, as indicated) are particularly important for staging (see Table 1, available online, in these guidelines, at www.NCCN.org [ST-1]) and planning therapy for oral cavity cancers, in addition to a complete H&N examination, biopsy, and other appropriate studies (see page 599). For patients who appear to have stage III/IV disease, PET-CT may alter management by upstaging patients.¹²⁵

Treatment

Surgery and radiotherapy represent the standards of care for early-stage and locally advanced resectable lesions in the oral cavity. The specific treatment is dictated by the TN stage and, if N0 at diagnosis, by the risk of nodal involvement (see pages 599–601). Multidisciplinary team involvement is particularly important for this site because of the critical physiologic functions of mastication, deglutition, and articulation of speech, which may be affected. Most panelists prefer surgical therapy for resectable oral cavity tumors, even for more advanced tumors. The concept of organ preservation using chemotherapy in the initial management of these patients has received less attention in the management of oral cavity cancers, because the functional outcome after primary surgical management is often good, given advances in reconstruction using microvascular techniques. Primary radiotherapy may be offered to select patients who are medically inoperable or refuse surgery.

Postoperative chemotherapy/radiotherapy (preferred, category 1) or reexcision of positive margins (if technically feasible) is recommended for all patients who have resected oral cavity cancers with the adverse pathologic features of extracapsular nodal spread and/or a positive mucosal margin.^61–64 For other risk features, such as pT3 or pT4 primary, N2 or N3 nodal disease, nodal disease in levels IV or V, or perineural invasion or vascular tumor embolism, clinical judgment should be used when considering whether to add chemotherapy to radiotherapy or treat with radiotherapy alone.

Follow-Up/Surveillance

Recommendations for surveillance are provided in the algorithm.

Workup and Staging

A multidisciplinary consultation is encouraged. Accurate staging (see Table 2, available online, at www.NCCN.org [ST-3]) depends on complete H&N examination coupled with appropriate imaging studies (see page 603).^12,126 Tumor HPV testing is suggested for cancers of the oropharynx given the established relationship between prior HPV infection and the development of a significant proportion of oropharyngeal cancers.

HPV Testing

Several studies have recently documented an increase in the incidence of HPV-related cancer, now estimated to constitute up to 60% to 70% of newly diagnosed cancers of the oropharynx in the United States and parts of the European Union.^11,127,128 A strong causal relationship has been established, particularly between HPV type 16 and the development of oropharyngeal cancer.⁴ Prospective and retrospective analyses of clinical trials indicate that patients with HPV-positive cancers have improved response to treatment and improved survival and progression-free survival when compared with HPV-negative tumors.^129–134 How this information should be used in routine clinical decision-making and in the design of clinical trials is currently a matter of intense investigation among NCCN Member Institutions. There is growing consensus that HPV status should be used as a stratification factor or should be addressed in separate trials (HPV related vs. unrelated disease) for which patients with oropharyngeal cancer are eligible.

Except for in cancers of unknown primary (see NCCN Clinical Practice Guidelines in Oncology for Occult Primary, available online, in these guidelines, at www.NCCN.org [OCC-1–6 and OCC-A]), the panel believes that HPV status currently should not be a routine consideration in treatment selection. Additional studies are needed to better understand the effect of HPV status on response to different therapies, treatment outcome, and patterns of failure, and in relation to other prognostic or predictive factors, such as smoking history and stage. Several clinical trial groups are reporting retrospective analyses of response to therapy in HPV-related versus HPV-unrelated oropharynx cancers.^{129–131,133} The panel strongly urges that patients with HPV-related cancers be enrolled in clinical trials evaluating biologic and treatment-related questions, when available.

HPV testing options in a clinical setting include HPV in situ hybridization [ISH]) and a surrogate marker, p16 immunohistochemistry (which is a more widely available test that has been shown in several studies to strongly correlate with HPV status and is similarly associated with improved prognosis).^{131–133,135} Sufficient pathologic material for HPV testing can be obtained by fine-needle aspiration (FNA).¹³⁶ The panel notes that HPV testing may prompt questions about prognosis (i.e., a favorable vs. less favorable forecast) and sexual history that the clinician should be prepared to address. Thus, without a specific reason for testing, HPV information may add anxiety and stress for some patients. Alternatively, understanding the cause of the cancer can result in reduced anxiety for some patients.

Treatment

The treatment algorithm has been divided into 3 staging categories: 1) T1–2, N0–1; 2) T3–4a, N0–1; and 3) any T, N2–3. Notably, T4b, any N, or unresectable nodal disease is treated as advanced cancer (see pages 616–618).

Early-stage (T1–2, N0–1) oropharyngeal cancers may be treated with primary surgery, including neck dissection, as indicated, or with definitive radiotherapy. The panel members felt that the third option of radiotherapy plus systemic therapy (category 2B for systemic therapy) was only appropriate for T2, N1 (see page 604). Adjuvant chemotherapy/radiotherapy is recommended (category 1) for adverse pathologic features of extracapsular nodal spread and/or positive mucosal margin.^61–63

For locally advanced resectable disease (T3–4a, N0–1; or any T, N2–3), 3 treatment approaches are recommended, in addition to enrollment in a multimodality clinical trial that includes function evaluation. The 3 approaches are: 1) concurrent systemic therapy/radiotherapy cisplatin (category 1; salvage surgery is used for managing residual or recurrent disease)⁹¹; 2) surgery with appropriate adjuvant therapy (chemo/radiotherapy or radiotherapy); or 3) induction chemotherapy followed by radiotherapy or chemo/radiotherapy, for which there was major disagreement among panel members.

Concurrent systemic therapy/radiotherapy with cisplatin alone (category 1) is preferred for treatment of locally or regionally advanced cancer (T3–4a, N0–1, or any T, N2–3) of the oropharynx. Panel members differed in their opinion as to whether induction chemotherapy should be considered a standard treatment option for T3–4a, N0–1 disease. This disagreement is reflected by a category 3 recommendation in the algorithms (see the following section on The Induction Chemotherapy Controversy).^91,137–146 Notably, for patients with any T, N2–3 disease, the category designation is 2B for induction chemotherapy because of the increased risk for distant metastases in patients with more advanced neck disease (see page 606).

The Induction Chemotherapy Controversy

Defining the optimal role of induction chemotherapy in the management of locally or regionally advanced H&N cancer has generated considerable discussion among the panel in recent years. The algorithm for the management of advanced oropharynx cancer illustrates well the lack of consensus among NCCN Member Institutions despite the extensive discussion. Thus, induction chemotherapy has a category 3 designation (major disagreement) for the management of T3–4a, N0–1 oropharyngeal disease. In addition, induction chemotherapy has a category 2B designation (non-uniform consensus, no major disagreement) for any T, N2–3 oropharyngeal disease. However, the lack of consensus is not unique to the oropharyngeal cancer algorithm; it is also apparent in other algorithms (see pages 608–613, 614 and 615, and 616–618) in which category 3 designations are common. Only for hypopharyngeal cancers less than T4a in extent (which, if managed surgically, would require total laryngectomy) is the use of induction chemotherapy—used here as part of a larynx preservation strategy—associated with a higher level of panel consensus (i.e., category 2A).

A brief review of the available data helps provide some perspective on the panel's deliberations. Most randomized trials of induction chemotherapy followed by radiotherapy and/or surgery compared with locoregional treatment alone published in the 1980s and 1990s did not show an improvement in overall survival with the incorporation of chemotherapy.¹⁴² However, a change in the pattern of failure with less-distant metastases was noted in some studies¹⁴⁷; also, a correlation seemed to be present between response to induction chemotherapy and subsequent durable response to radiation.^147,148 Thus, the concept developed that in selected patients, induction chemotherapy could facilitate organ preservation, avoid morbid surgery, and improve overall quality of life even though overall survival was not improved. Because total laryngectomy is among the procedures most feared by patients,¹⁴⁹ larynx preservation was the focus of initial studies.

Two randomized studies, the Veterans Affairs (VA) Laryngeal Cancer Study Group trial in advanced larynx cancer and the EORTC trial predominantly in advanced hypopharynx cancer, established the role of induction cisplatin/5-FU chemotherapy followed by definitive radiotherapy in responding patients as an alternative treatment to primary total laryngectomy and postoperative radiation, offering potential larynx preservation without compromise in survival (see page 639).^147,148 Yet even in this setting, the role of induction chemotherapy decreased with time. Randomized trials and related meta-analyses indicated that concurrent chemoradiotherapy (with cisplatin being the best-studied agent) offered superior locoregional tumor control and survival over radiation alone,^150–160 and shorter duration of therapy than induction therapy followed by radiation. Meta-analyses reported that concurrent chemoradiotherapy was more efficacious than an induction chemotherapy strategy.^142,146 In the larynx preservation setting, Intergroup 91-11 compared radiation alone, concurrent cisplatin/radiation, and induction cisplatin/5-FU followed by radiation, all with surgery for salvage. The concurrent arm had the highest larynx preservation rate (see Cancer of the Larynx, in these guidelines, online, at www.NCCN.org [MS-19]).¹⁶¹

Nonetheless, renewed interest has been shown in the role of induction chemotherapy for a few reasons. Given improvements in locoregional control now achieved with advances in surgery, radiotherapy, and concurrent chemotherapy/radiotherapy, the role of distant metastases as a source of treatment failure has increased and induction chemotherapy allows greater drug delivery for this purpose.¹⁶² Concern has been growing regarding the long-term morbidity of concurrent chemoradiotherapy and related increasing interest in exploring alternative approaches that might have a different and hopefully more favorable side effect profile. Finally, a more effective triplet chemotherapy regimen has been identified compared to the standard cisplatin/5-FU used in induction trials of the 1980s and 1990s, and the related meta-analyses. Results from 3 phase III trials, which compared induction cisplatin plus infusional 5-FU with or without the addition of a taxane (docetaxel or paclitaxel) followed by the same locoregional treatment, showed significantly improved outcomes (response rates, disease-free survival, or overall survival depending on the trial) for patients in the 3-drug induction group compared with those receiving 2 drugs (cisplatin plus 5-FU).^{139,141,144,145} A randomized trial in the larynx preservation setting similarly showed superior larynx preservation outcomes when induction docetaxel/cisplatin/5-FU (TPF) and cisplatin/5-FU were compared.¹⁶³

However, a clear advantage in overall survival from the addition of induction chemotherapy to concurrent chemoradiation has not been shown yet. A randomized phase II study in patients with stage III or IV squamous cell H&N cancer of induction TPF followed by concurrent cisplatin/5-FU with radiotherapy versus concurrent cisplatin/5-FU with radiotherapy alone did report a higher radiologic complete response rate with the incorporation of induction chemotherapy.¹⁶⁴ However, a randomized 3-arm study comparing concurrent cisplatin/radiotherapy versus induction chemotherapy with TPF or cisplatin/5-FU followed by concurrent cisplatin/radiotherapy reported a decrease in time to treatment failure with the incorporation of induction therapy, but no difference in survival. Furthermore, approximately 3 times as many patients were not included in the efficacy assessments on the induction arms, suggesting potential toxicity concerns.¹⁶⁵

There also remains considerable uncertainty and disagreement among panel members concerning which radiation or chemoradiation plan should follow induction.¹⁶⁶ Panel members agree that high-dose cisplatin (100 mg/m² every 21 days × 3) may not be feasible for many patients in this setting,^165,167 raising concerns that any efficacy gains of induction may be offset by the use of better-tolerated but potentially less-effective concurrent programs or poorer patient compliance with the radiation-based part of treatment. No one concurrent chemotherapy regimen is preferred. Panel members agreed that many different alternatives are reasonable (including concurrent low-dose weekly cisplatin, weekly taxanes, cetuximab, or combinations thereof), but are inadequately studied, to be specifically recommended.¹⁶⁸

After induction chemotherapy, the use of cetuximab is supported by data from the TREMPLIN study, in which patients with advanced laryngeal or hypopharyngeal cancer who had a major response to induction TPF were randomized to high-dose cisplatin for 3 cycles versus weekly cetuximab concurrent with radiotherapy. Patients on the cetuximab arm tolerated therapy better, had better compliance with drug delivery, and had similar 3-month laryngeal preservation rates to those observed in patients on the cisplatin arm.¹⁶⁷ Some panel members specifically considered exclusive use of low-dose weekly carboplatin in this setting to be inadequate.¹⁶⁹ Some evidence suggests that chemotherapy with weekly carboplatin might be equivalent to cisplatin; however, data are from the nasopharyngeal setting.¹⁷⁰ Other sequential induction-concurrent regimens, using less-aggressive induction or less-intensive concurrent chemotherapy, seem to have higher compliance rates.^141,171 However, a definitive study has not compared these newer strategies with concurrent chemoradiotherapy alone.

Because of these uncertainties, enrolling patients in appropriate clinical trials is particularly encouraged. Outside of a clinical trial, proceeding directly to concurrent chemoradiotherapy, cisplatin preferred, remains a standard treatment option for these patients, and is the preferred approach from the panel's perspective in several settings as indicated. When induction chemotherapy is used, randomized data have clearly proven that the addition of a taxane to cisplatin/5-FU, of which TPF is the most extensively studied, is more efficacious than cisplatin/5-FU.

Radiation Therapy Fractionation

Standard conventional fractionation is preferred when radiotherapy is used definitively for T1–2, N0 tumors. Altered fractionation is appropriate for selected T1–2, N1 tumors, particularly if concurrent chemotherapy is not used. The recommended schedules are shown in the algorithm (see pages 603–607).

Follow-Up/Surveillance

Recommendations for surveillance are provided in the algorithm.

Workup and Staging

A multidisciplinary consultation is encouraged. Accurate staging (see Table 2, available online, in these guidelines, at www.NCCN.org [ST-3]) depends on a complete H&N examination coupled with appropriate studies (see page 608).¹²

At diagnosis, approximately 60% of patients with cancer of the hypopharynx have locally advanced disease with spread to regional nodes. Furthermore, autopsy series have shown a high rate of distant metastases (60%) involving virtually every organ.¹⁷² Thus, the prognosis for patients with cancer of the hypopharynx can be poor despite aggressive combined modality treatment.

Treatment

Patients with resectable disease are divided into 2 groups: 1) those with early-stage cancer (most T1, N0; selected T2, N0) who do not require a total laryngectomy; and 2) those with advanced resectable cancer (T1, N+; T2–4a, any N) who do require laryngectomy. The surgery and radiotherapy options for the former group (see page 610) represent a consensus among the panel members.

Patients with more advanced disease (defined as T1, N+; T2–3, any N) requiring total laryngectomy and partial or total pharyngectomy may be managed with 3 approaches (see page 610) in addition to enrollment in multimodality clinical trials: 1) induction chemotherapy followed by definitive radiotherapy if a complete response was achieved at the primary site¹⁴⁷ or followed by other options depending on the response (see page 611); 2) surgery with neck dissection and postoperative radiation or chemoradiation as dictated by pathologic risk features; or 3) concurrent chemotherapy/radiotherapy, cisplatin preferred. Given the functional loss resulting from this surgery and the poor prognosis, participation in clinical trials is emphasized.

The recommendation of the induction chemotherapy/definitive radiotherapy option is based on the results of an EORTC randomized trial.¹⁴⁷ This trial enrolled 194 eligible patients with stage II, III, or IV resectable squamous cell carcinoma of the pyriform sinus (n = 152) and aryepiglottic fold (n = 42), excluding patients with T1 or N2c disease. Patients were randomly assigned either to laryngopharyngectomy and postoperative radiotherapy, or to chemotherapy with cisplatin and 5-FU for a maximum of 3 cycles, followed by definitive radiotherapy. A complete response to induction chemotherapy was required to proceed with definitive radiotherapy. The published results showed equivalent survival, with median survival duration and 3-year survival rate of 25 months and 43%, respectively, for the surgery group versus 44 months and 57%, respectively, for the induction chemotherapy group.¹⁴⁷ A functioning larynx was preserved in 42% of patients who did not undergo surgery. Local or regional failure rates did not differ between patients treated with surgery and those treated with chemotherapy, although the chemotherapy recipients showed a significant reduction in distant metastases as a site of first failure (P = .041). Adherence to the requirements for complete response to chemotherapy and for inclusion of only patients with the specified TN stage is emphasized.

A recently published randomized trial showed that an alternating program of cisplatin/5-FU with radiotherapy yielded larynx preservation, progression-free interval, and overall survival rates equivalent to those obtained with induction platinum/5-FU followed by radiotherapy.¹⁷³ Given available randomized data showing the superiority of TPF compared with PF for induction chemoradiation, the triplet is now recommended as induction for this approach.^149,154,163

As noted in the algorithm, surgery is recommended if less than a partial response occurs after induction chemotherapy (see page 611). In this situation, or when primary surgery is the selected management path, postoperative chemotherapy/radiotherapy is recommended (category 1) for the adverse pathologic features of extracapsular nodal spread and/or positive mucosal margin. For other risk features, clinical judgment should be used when deciding to use radiotherapy alone or when considering adding chemotherapy to radiotherapy.

Options for patients with T4a, any N disease (see page 612) include surgery plus neck dissection (preferred) followed by adjuvant chemotherapy/radiotherapy or radiotherapy; multimodality clinical trials; or category 3 recommendations.

Follow-Up/Surveillance

Recommendations for surveillance are provided in the algorithm.

Workup and Staging

The workup of nasopharyngeal cancer includes a complete H&N examination and other studies (see page 614), which are important for determining the full extent of tumor to assign appropriate stage and to design radiation ports that will encompass all disease with appropriate doses. Multidisciplinary consultation is encouraged. The 2010 AJCC staging classification (7th edition) is used as the basis for treatment recommendations (see Table 2, available online, in these guidelines, at www.NCCN.org [ST-3]).¹²

Treatment

Patients with T1, N0, M0 nasopharyngeal tumors may be treated with definitive radiotherapy alone (see page 615). For early-stage cancer in this setting, radiation doses of 66 to 70 Gy given with standard fractions are necessary for control of gross tumor (see page 615). The local control rate for these tumors ranges from 80% to 90%, whereas T3–4 tumors have a control rate of 30% to 65% with radiotherapy alone.^180,181

The combination of radiotherapy and concurrent platinum-based chemotherapy followed by adjuvant cisplatin/5-FU has been shown to increase the local control rate from 54% to 78%. The Intergroup trial 0099, which randomly assigned patients to chemotherapy plus external-beam radiotherapy versus external radiation alone, closed early when an interim analysis disclosed a significant survival and progression-free survival advantage favoring the combined chemotherapy and radiation group.¹⁷⁹ The addition of chemotherapy also decreased local, regional, and distant recurrence rates.

A similar randomized study conducted in Singapore, which was modeled after the Intergroup treatment regimen, continued to show the benefit of the addition of chemotherapy to radiotherapy. Adjuvant chemotherapy after combined chemotherapy and radiation was also given in this trial.¹⁸² In addition, the administration of the cisplatin dose was spread out over several days, and this regimen seemed to reduce toxicity while still providing a beneficial antitumor effect.

Another phase III randomized trial showed that concurrent chemo/radiotherapy (using weekly cisplatin) increased survival compared with radiotherapy alone.¹⁸³ The 5-year overall survival was 70% for the chemo/radiotherapy group versus 59% for the radiotherapy group. A randomized trial compared chemo/radiotherapy using cisplatin versus carboplatin and found that the 3-year overall survival rates were similar (78% vs. 79%).¹⁷⁰ However, these NCCN Guidelines recommend cisplatin for chemo/radiotherapy in patients who do not have a contraindication to the drug, because more randomized data support the use of cisplatin in this setting.

These NCCN Guidelines recommend concurrent chemotherapy (cisplatin) plus radiotherapy (category 1) followed by adjuvant cisplatin/5-FU for T1, N1–3; and for T2–T4, any N lesions (see page 615). Although an unusual occurrence, patients with residual disease in the neck and a complete response at the primary should undergo a neck dissection. Initial therapy for patients who present with metastatic disease should consist of a platinum-based combination chemotherapy regimen (see page 615).

The management of patients with recurrent or persistent nasopharyngeal cancer is described on page 617. When chemotherapy is indicated, commonly used active agents alone or in combination include gemcitabine, paclitaxel, docetaxel, cisplatin, or carboplatin.^184–188 Cetuximab plus carboplatin has been studied for patients with recurrent or metastatic nasopharyngeal cancer for whom platinum-based therapy has failed¹⁸⁵; however, this regimen is not currently recommended in these NCCN Guidelines.

Follow-Up/Surveillance

Recommendations for surveillance are provided in the algorithm.

Treatment

Participation in clinical trials is preferred for all patients with very advanced H&N cancers.

Newly Diagnosed Advanced Disease: For patients with a performance status (PS) of 0 or 1, the standard treatment of newly diagnosed, very advanced disease is concurrent cisplatin chemotherapy and radiotherapy (category 1).¹⁵⁰ The panel had a major disagreement regarding whether induction chemotherapy (TPF) followed by radiotherapy or chemoradiation should be used for patients with a PS of 0 or 1, which is reflected in the category 3 recommendation (see The Induction Chemotherapy Controversy, page 637).^141,145 Other options for patients with PS 2 or 3 are described in the algorithm (see page 616).

Many randomized trials^{64,88,89,150–156} and meta-analyses of clinical trials^{142,157–160} show significantly improved overall survival, disease-free survival, and local control when a concomitant or alternating chemotherapy and radiation regimen is compared with radiotherapy alone for advanced disease. All combined chemoradiotherapy regimens are associated with various degrees of enhanced mucosal toxicities, which require close patient monitoring, ideally provided by a team experienced in treating patients with H&N cancer. Limited data are available comparing the efficacy of different chemoradiotherapy regimens. Single-agent cisplatin plus radiotherapy is effective and easy to administer, and typically uses conventional fractionation at 2.0 Gy/fraction to 70 Gy or more in 7 weeks, with single-agent cisplatin given every 3 weeks at 100 mg/m² (see page 618).¹⁵⁰

Bonner et al.¹⁸⁹ randomly assigned 424 patients with locally advanced and measurable stage III/IV squamous cell carcinomas of the H&N to receive definitive radiotherapy with or without cetuximab. Locoregional control and median overall survival (49 vs. 29.3 months; P = .03) were significantly improved in patients treated with radiotherapy and cetuximab compared with radiotherapy alone. Radiotherapy and cetuximab may provide a therapeutic option for patients not considered medically fit for standard chemoradiotherapy regimens.

Other preferred chemoradiation options were also identified by the panel (see page 626).^190,191 Limited data are available comparing combination chemoradiation with a single-agent used concurrently with radiotherapy.

Recurrent or Persistent Disease: Surgery is recommended for resectable recurrent or persistent locoregional disease; adjuvant therapy depends on the risk factors (see page 617). If the recurrence is unresectable and the patient did not have prior radiotherapy, then radiotherapy with concurrent systemic therapy is recommended, depending on the PS (see page 617). The treatment approach for patients with recurrent disease not amenable to curative-intent radiation or surgery is the same as that for patients with metastatic disease; enrollment in a clinical trial is preferred.

Metastatic Disease: Palliative adjunctive measures include radiotherapy to areas of symptomatic disease, analgesics, and other measures to control other manifestations of disease spread (e.g., hypercalcemia). Single agents and combination systemic chemotherapy regimens are both used (see page 626). Response rates to single agents range from 15% to 35%. The most active single agents include cisplatin, carboplatin, paclitaxel, docetaxel, 5-FU, methotrexate, ifosfamide, bleomycin, gemcitabine (for nasopharyngeal cancer), and cetuximab (for non-nasopharyngeal cancer).^{186,192–194} Active combination regimens include 1) cisplatin or carboplatin, plus 5-FU^195,196 with cetuximab (for non-nasopharyngeal cancer)¹⁹⁷; 2) cisplatin or carboplatin, plus a taxane^195,198; 3) cisplatin with cetuximab (for non-nasopharyngeal cancer)¹⁹⁹; or 4) cisplatin with 5-FU.^195,196 These regimens, on average, result in a doubling of response rates compared with single agents.

Randomized trials assessing a cisplatin-based combination regimen (e.g., cisplatin plus 5-FU) versus single-agent therapy with cisplatin, 5-FU, or methotrexate have shown significantly higher response rates for the combination regimen. No difference in overall survival has been seen.^{195,196,200–202} Historically, the median survival with chemotherapy is approximately 6 months, and the 1-year survival rate is approximately 20%. Achievement of a complete response is associated with longer survival and, although infrequent, has been reported more often with combination regimens.¹⁹⁶ A randomized phase III trial comparing cisplatin plus 5-FU with cisplatin plus paclitaxel in patients with metastatic or recurrent H&N cancer found no significant differences in survival.¹⁹⁵

The epidermal growth factor receptor (EGFR) is a trans-membrane glycoprotein; activation of EGFR triggers a cascade of downstream intracellular signaling events important for regulation of epithelial cell growth. Overexpression of EGFR and/or common ligands has been observed in greater than 90% of H&N squamous cell carcinomas. This finding has led to the development of EGFR inhibitors, such as the monoclonal antibody cetuximab and small molecule tyrosine kinase inhibitors (i.e., erlotinib and gefitinib).

Data from phase II studies indicate that in the cisplatin-refractory setting, the single-agent response rate of cetuximab is 12% to 14%. Burtness et al.¹⁹⁹ compared cisplatin plus cetuximab versus cisplatin plus placebo as first-line treatment of recurrent disease and reported a significant improvement in response rate with cetuximab (26% vs. 10%, respectively). Notably, a phase III randomized trial (EXTREME) of 442 patients with recurrent or metastatic squamous cell carcinoma found that cetuximab plus cisplatin/5-FU or carboplatin/5-FU improved median survival compared with the standard chemotherapy doublet (10.1 vs. 7.4 months; P = .04).¹⁹⁷ The response rate was also improved with cetuximab (20%–36%; P < .001). Available data for tyrosine kinase inhibitors (such as erlotinib and gefitinib) have not established them as treatment options for recurrent or metastatic H&N cancer outside of a clinical trial. In one randomized trial, treatment with 2 different dosing schedules of gefitinib offered no survival advantage over treatment with methotrexate.²⁰³

The standard treatment of patients with incurable, recurrent, or metastatic H&N cancer should be largely dictated by the patient's PS (see page 616). Patients should be fully informed about the goals of treatment, cost of combination chemotherapy, and potential for added toxicity.

Individual Disclosures for the NCCN Guidelines Panel for Head and Neck Cancers

View Table