Predictive Accuracy of First Post-Treatment PET/CT in HPV-Related Oropharyngeal Squamous Cell Carcinoma

Transcription

1 The Laryngoscope VC 2014 The American Laryngological, Rhinological and Otological Society, Inc. Predictive Accuracy of First Post-Treatment in HPV-Related Oropharyngeal Squamous Cell Carcinoma Yekaterina Koshkareva, MD; Barton F. Branstetter IV, MD; John P. Gaughan, PhD; Robert L. Ferris, MD, PhD Objectives/Hypothesis: To determine whether the result of first posttreatment positron emission tomography and computed tomography () is predictive of outcome in patients with oropharyngeal squamous cell carcinoma (OPSCC), and whether accuracy is affected by human papillomavirus (HPV) status. Study Design: Retrospective review. Methods: Demographic, clinical, and radiographic data were available for 61 patients with OPSCC, treated in 2004 to 2012 at a single tertiary academic referral center, with at least one baseline and one posttreatment. Clinical followup was obtained every 3 months thereafter. The median follow-up time was 36 months (range months). Results: Of 61 patients, 48 (79%) had negative first posttreatment results; and overall, 18 of the 61 patients (30%) recurred. All accuracy measures for were higher in HPV-positive patients, including a 93% negative predictive value (NPV). Patients with positive results had poorer survival on Kaplan-Meier analyses. On multivariate analysis of factors predictive of recurrence, two parameters were significant: HPV status (P ) and result (P <0.0001). Conclusions: A negative first posttreatment result is associated with better prognosis and rare recurrence, especially in patients with HPV-positive status. Less frequent radiologic surveillance is warranted in patients with HPVpositive OPSCC and a negative first posttreatment scan. Key Words:, head and neck squamous cell carcinoma (HNSCC), oropharynx, human papillomavirus, surveillance. Level of Evidence: 2b. Laryngoscope, 124: , 2014 INTRODUCTION According to the National Cancer Institute, there will be 13,930 new cases of pharyngeal cancers in the United States in 2013, and over 2,000 Americans will die from it. 1 There has been a significant increase in human papillomavirus (HPV)-related oropharyngeal squamous cell carcinoma (OPSSC) of the base of tongue and tonsil since the 1970s. 2 In fact, some reports project that if the recent incidence patterns continue, the annual number of HPV-related OPSCC will approximate or even surpass the annual number of cervical cancers by the year HPV-related cancers tend to affect younger patients, present with smaller primary tumors, and predominantly involve the oropharynx. HPV-related cancers respond to treatment better, have longer overall From the Department of Otolaryngology (Y.K., B.F.B., R.L.F.); the Department of Radiology, Division of Neuroradiology (B.F.B.), University of Pittsburgh School of Medicine, Pittsburgh; and the Biostatistics Consulting Center (J.P.P.), Temple University School of Medicine, Philadelphia, Pennsylvania, U.S.A. Editor s Note: This Manuscript was accepted for publication January 23, Presented as a poster presentation at Combined Otolaryngology Spring Meeting, Orlando, Florida, U.S.A., April 10, The authors have no funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Robert L. Ferris, MD, PhD, Department of Otolaryngology, Eye and Ear Institute, University of Pittsburgh School of Medicine, 200 Lothrop St, Suite 500, Pittsburgh, PA ferrrl@upmc.edu DOI: /lary and progression-free survival, and have lower locoregional recurrence rates. 4,5 Despite many available treatment modalities, historically almost half of OPSSC patients have suffered recurrence, even after treatment with curative intent. 6 Almost all recurrences arise in the first 3 years of treatment. In the absence of clinical signs suggestive of recurrence, such as increasing pain, dysphagia, or new ulceration or mass, a physician must rely on clinical or endoscopic evaluation or imaging techniques such as computed tomography (CT) and magnetic resonance imaging. Although endoscopic evaluation with tissue sampling remains the gold standard for diagnosis, this technique is invasive and may result in morbidity or even mortality. Imaging is the preferred method for surveillance. Both radiation and surgery distort normal anatomy, violate tissue planes, and alter anatomic landmarks, making both clinical and cross-sectional imaging evaluations challenging. One of the functional imaging techniques useful for recurrence detection is positron emission tomography (PET). It utilizes 18F-fluorodeoxyglucose (18F-FDG), a radioactively labeled positron emitting tracer of glucose uptake. A systematic review by Isles et al. report 94% sensitivity for detection of residual or recurrent disease in head and neck squamous carcinoma. 7 Multiple reports provide evidence of even greater accuracy with combined PET and CT () In fact, is now included in the National Comprehensive Cancer Network Clinical Practice Guidelines as 1843

2 a supplement to clinical evaluation in the posttreatment surveillance of patients with cancers of oral cavity, oropharynx, hypopharynx, larynx, and unknown primary. Also, 95% of clinically asymptomatic recurrences can be detected within the first 24 months when is used for surveillance. 11 No evidence-based guidelines exist on timing and duration of surveillance. It has been widely reported that a 3-month posttreatment is able to distinguish persistent disease from treatment-related changes with a high negative predictive value. 12,13 However, is a costly technique with substantial radiation exposure. 14,15 Hence, the benefit of continuous routine surveillance remains uncertain. Recently, Zhang et al. demonstrated that the prognostic value of in head and neck cancer (HNC) patients may be affected by the HPV status of the tumor. 16 The purpose of this study is to evaluate the predictive value of the first posttreatment with respect to disease progression and a 2-year disease-free survival in the specific patient population of patients with OPSCC. We attempted to stratify our results by the HPV status of the tumors. immediately preceding PET. The PET images were then reconstructed with CT-based attenuation correction. The axial CT images were reviewed on a conventional picture archiving and communication system workstation (Philips isite, Andover, MA). fusion images, including multiplanar reformatted images, were reviewed on dedicated workstation (Vital Images, Minnetonka, MN). All images were interpreted by a dedicated head and neck radiologist with certificate of added qualification certification in neuroradiology and at least 3 years of experience with of head and neck squamous cell carcinoma (HNSCC). Interpretation was performed as a normal part of clinical care, with all clinical information available to the interpreting physician. Strict standard uptake values (SUV) thresholds were not utilized, in keeping with the usual practice pattern at our institution. The interpretations were classified into one of the four categories: 1) negative; 2) probably negative (mild residual FDG uptake, or near-complete tumor regression); 3) probably positive (moderate residual FDG uptake concerning for tumor persistence); or 4) positive (increased FDG uptake or suspicious radiographic characteristics suggestive of tumor persistence or definite tumor progression at the primary site, in cervical lymph nodes, or distant sites). When calculating accuracy values, negative and probably negative categories were combined, and probably positive and positive categories were combined. All four categories were used when calculating Kaplan-Meier survival curves. MATERIALS AND METHODS After approval from the institutional review board at the University of Pittsburgh, a retrospective review was conducted in compliance with the Health Information Portability and Accountability Act. Patients Our tumor registry data was reviewed for all patients with biopsy-proven OPSCC who were evaluated at our institution between January 2004 and September Sixty-one previously untreated patients with known HPV status (tested at the University of Pittsburgh Medical Center), who were treated with curative intent and followed with combined imaging for posttreatment surveillance, were included in the study. The following exclusion criteria were established: 1) distant metastatic disease at the time of presentation; 2) previous history of head and neck malignancy; 3) history of previous radiation to the head and neck region, and 4) performed for purposes other than routine posttreatment surveillance, such as evaluation of a concerning clinical finding. FDG- Imaging and Interpretation According to the posttreatment surveillance protocol at the University of Pittsburgh Medical Center (UPMC), first posttreatment is performed 8 weeks after completion of treatment. 17 Combined imaging was performed on a Reveal scanner (CTI Medical System, Knoxville, TN). A dual-channel CT was combined with a lutetium oxyorthosilicate Allegra PET. Patients were instructed to fast at least 4 hours prior to acquisition. Blood glucose levels were measured before 18 F-Fluorodeoxyglucose (FDG) injections; patients whose blood glucose levels were over 200 mg/dl were rescheduled. imaging of the head, neck, chest and abdomen was performed 60 minutes after an intravenous injection of 8 15 mci FDG. Helical CT (pitch 5 1.0; mas ; kvp 130) with intravenous contrast (125 ml Optiray-350) was performed 1844 Posttreatment Surveillance Patients were evaluated by their treating otolaryngologist 1 month after completion of their treatment. At that time, the first posttreatment was ordered according to UPMC HNSCC surveillance protocol as described above. Patients were then clinically and radiologically evaluated every 3 months, supplemented by additional imaging, examinations under anesthesia, or tissue sampling as appropriate. Human Papillomavirus Detection The paraffin-embedded tumor tissue from each patient was subjected to in situ hybridization for detection of HPV. The immunohistochemical analysis determined the p16 status. (Positive p16 expression was defined as strong and diffuse nuclear and cytoplasmic staining in at least 70% of the tumor cells.) Statistical Analysis The primary goal of this study was to determine the accuracy of the first posttreatment to predict 2-year disease-free survival (DFS). DFS and follow-up period were calculated from the date of treatment completion to the date of recurrence detection or most recent follow-up. Disease-free survival was evaluated using Kaplan-Meier method. Recurrence within 2 years was used as the reference for calculation of predictive value of. Sensitivity, specificity, negative predictive value (NPV), positive predictive value, and overall accuracy were calculated. The predictive values of the first reading, HPV status, and other clinical covariates (age, gender, tumor [T] and node [N] stages and treatment) were analyzed using Fisher s exact test. Those parameters that were significant in the univariate analysis were then included in a multivariate model using stepwise elimination. RESULTS Using our prospectively collected tumor registry, 70 patients with histologically proven OPSCC and known

3 TABLE I. Demographics and Clinical Information (N 5 61). Characteristic Total (%) Probably Probably Negative Negative Positive Positive N 5 31 (51%) N 5 17 (28%) N 5 4 (6%) N 5 9 (15%) Age average range gender male 52 (85) female 10 (15) T stage stage 1 14 (23) stage 2 33 (54) stage 3 9 (15) stage 4 5 (8) N stage stage 0 7 (11.5) stage 1 13 (21.3) stage 2 39 (64) stage 3 2 (3.2) Disease site tonsil 31 (51) base of tongue 29 (48) posterior pharynx 1 (1) HPV status (%) HPV Positive 50 (82) HPV Negative 11 (18) HPV 5 human papillomavirus; 5 posttreatment positron emission tomography and computed tomography. HPV status were treated with curative intent at our institution from 2004 to October Nine patients were excluded due to insufficient follow-up or missing clinical data, and the remaining 61 patients were included in the analysis. The mean age was 52 years (range 36 69); 52 (82%) were males. The primary tumor site was tonsil in 31 (51%) patients, base of tongue in 29 (48%) patients, and posterior pharyngeal wall in 1 (1%) patients. Eighty-two percent (82%) of tumors were HPVpositive. Patient demographic information and tumor characteristics are summarized in Table I. For the purpose of the survival curves, the median follow-up time was 36 months (range months). The average time from completion of treatment to first posttreatment was 9 weeks (range 6 15 weeks). There were 18 recurrences among the 61 patients, 60.7% of HPV1 patients were smokers. The overall rate of recurrence in HPV-positive patients was 24% (12/50); the recurrence rate in HPV-negative patients was 55% (6/11). The predictive accuracy of overall is tabulated in Table II; the values for HPV-positive patients are shown in Table III. All of the accuracy values were superior in HPV-positive patients. Most notably, the NPV rose from 85% in the overall population to 93% in the HPV-positive cohort. On Kaplan-Meier analysis, negative and probably negative curves were not substantially different, but patients with probably positive and positive results fared significantly worse (Fig. 1). Univariate analysis revealed that HPV status (P ), T stage (P ), and category (P < ) were predictive of recurrence. On multivariate analysis, however, only two parameters remained significant: HPV status (P ) and result (P < ). TABLE II. Predictive Accuracy of First Posttreatment. Disease Status at 2 Years Outcome NED Recurrence Total Negative Positive Total NPV 5 85%; PPV 5 84%; sensitivity 5 61%; specificity 5 95%. NED 5 no evidence of disease; NPV 5 negative predictive value; 5 posttreatment positron emission tomography and computed tomography; PPV 5 positive predictive value. 1845

4 TABLE III. Predictive Accuracy of in HPV1 Patients. Disease Status at 2 Years Outcome NED Recurrence Total Negative Positive Total NPV 5 93%; PPV 5 90%; sensitivity 5 75%; specificity 5 97%. HPV 5 human papillomavirus; NED 5 no evidence of disease; NPV 5 negative predictive value; 5 posttreatment positron emission tomography and computed tomography; PPV 5 positive predictive value. DISCUSSION This study is the first report of an excellent NPV (93%) from the first posttreatment in a targeted HPV-positive OPSCC patient population. Both HPV status and negative reading were more predictive of disease-free survival than tumor stage in patients with OPSCC. Combined imaging is now included in the National Comprehensive Cancer Network Clinical Practice Guidelines as a supplement to clinical evaluation in posttreatment surveillance of patients with OPSCC. 1 Results of the first posttreatment were found predictive of the outcome in HNC patients. 13 Furthermore, Zhang et al. demonstrated that the prognostic value of the negative first posttreatment is even higher in HPV-positive patients. However, multiple disease sites were included and few HPV-positive tumors supported the major conclusions of that report (n 5 14). 16 The majority of the HPV-related head and neck tumors are oropharyngeal. Hence, we focused our analysis on patients with OPSCC. Fifty patients (82%) had HPV-related OPSCC tumors in accordance with recently expected frequencies. 18 As expected, first negative and probably negative readings were associated with better 2-year disease-free survival. The NPV of a negative at 2 years was 85%. This value was further augmented by positive HPV status, raising the NPV to 93%. The univariate analysis identified PET/ CT reading, HPV status, and T stage as the predictors of a favorable outcome, but on multivariate analysis only and HPV status remained significant. This may be explained by the distribution of tumors in our cohort: Approximately one-half of our patients had stage T2 tumors, and only one-quarter of our patients had stage T3 and T4 tumors. These findings suggest that patients with HPVrelated OPSCC and negative may benefit from a less intense radiographic surveillance, reducing radiation exposure and health care costs. This suggestion should be directly tested in prospective trials, randomizing patients to CT alone versus to assess the impact on survival. There have been a number of studies demonstrating the high accuracy and NPV of in head and neck cancer surveillance. 9,13,16,19 22 Abgral et al. reported accuracy of 90% and NPV of 100% in 99 patients. The survival benefit, however, was not assessed. 9 Kao et al. again reported NPV of 98% and demonstrated significantly better 2-year progression-free and 2-year overall survival rates in patients with negative within 6 months of radiation completion (review of 80 patients). 13 Similarly, Yao et al. reported PET NPV of 99% and demonstrated better 3-year overall and diseasefree survival in PET negative cohort in their analysis of 171 patients. They also proposed the use of postradiation PET to individualize the surveillance plans, making them less vigorous in patients with negative PET results. 19 Ho et al. questioned the benefit of additional 12- and 24-month screening beyond 3 months. The 12-month identified clinically occult disease in 9% of patients and 24-month identified clinically occult disease in 4% of patients. However, there were statistically significant differences in median disease-free interval or overall survival between and clinically detected recurrences. 21 A prospective analysis of 98 patients by Moeller et al. showed that was more accurate than CT alone in the assessment of treatment response in patients at high risk for treatment failure (defined as tobacco-related, HPV-negative, and non-op primary tumors). There was no benefit of addition of in low-risk patients. 20 The review of 62 patients by Zhang et al. out of our institution confirmed improved better disease-free survival in patients with negative first posttreatment readings and HPV-related tumors. 16 Our study is the largest review of in head and neck cancers limited to an oropharyngeal location. The high NPV of and improved survival in patients with HPV-related tumors and negative first posttreatment readings are in agreement with the existing literature. There are four major limitations of this study. First, any retrospective studies are limited by the availability and heterogeneity of data. Second, the limitation is the selection bias of the patient population in a tertiary care center. The patients seen at high-volume tertiary care institutions may not be representative of the general population. Third, the radiologic evaluation is uniform and standardized according to the surveillance protocol within the radiology department of our institution only; it may not apply in situations for which different surveillance protocols Fig. 1. Kaplan-Meier estimates of recurrence stratified by the result of the first posttreatment (circles represent censored values). 5 posttreatment positron emission tomography and computed tomography. [Color figure can be viewed in the online issue, which is available at

5 are employed. Finally, the small number of HPV-negative patients in our cohort renders the comparison between HPV-positive and HPV-negative patients less robust. CONCLUSION A negative first posttreatment is associated with better prognosis and rare recurrence, especially in HPV-related OPSCC patients. An alternative, less frequent surveillance scheme may be more cost effective for this patient population. Acknowledgement We wish to thank Jennifer Hetrick, RHIA, CRT, for her assistance in UPMC Network Cancer Registry Head and Neck Cancer database data extraction. The work was done at the University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania, and funded internally by the Departments of Otolaryngology and Radiology. BIBLIOGRAPHY 1..NCCN clinical practice guidelines in oncology: head and neck cancers. National Comprehensive Cancer Network Web site. professionals/physician_gls/pdf/head-and-neck.pdf. Accessed February 21, Chatuvedi AK, Engels EA, Anderson WF, Gillison ML. Incidence trends for human papillomavirus-related and unrelated oral squamous cell carcinomas in the United States. J Clin Oncol 2008;26: Chatuvedi AK, Engels EA, Pfeiffer RM, et al. Human papillomavirus and rising oropharyngeal cancer incidence in the United States. J Clin Oncol 2011;29: Fakhry C, Westra WH, Li S, et al. Improved survival of patients with human papillomavirus-positive head and neck squamous cell carcinoma in a prospective clinical trial. J Natl Cancer Inst 2008;100: Ang KK, Harris J, Wheeler R, et al. Human papillomavirus and survival of patients with oropharyngeal cancer. N Engl J Med 2010;363: Ang KK, Trotti A, Brown BW, et al. Randomized trial addressing risk features and time factors of surgery plus radiotherapy in advanced head and neck cancer. Int J Radiat Oncol Biol Phys 2001;51: Isles MG, McConkey C, Mehanna HM. A systematic review and metaanalysis of the role of positron emission tomography in the follow up of head and neck squamous cell carcinoma following radiotherapy or chemoradiotherapy. Clin Otolaryngol 2008;33: Agarwal V, Branstetter BF, Johnson JT. Indications for in the head and neck. Otolaryngol Clin North Am 2008;41: Abgral R, Querellou S, Potard G, et al. Does 18F-FDG improve the detection of post-treatment recurrence of head and neck squamous cell carcinoma in patients negative for disease on clinical follow-up? J Nucl Med 2009;50: Branstetter BF, Blodgett TM, Zimmer LA, et al. Head and neck malignancy: is more accurate that PET or CT alone? Radiology 2005; 235: Beswick DM, Gooding WE, Johnson JT, Branstetter BF. Temporal patterns of head and neck squamous cell carcinoma recurrence with positronemission tomography computed tomography monitoring. Laryngoscope 2012;122: Gupta T, Master Z, Kannan S, et al. Diagnostic performance of posttreatment FDG PET or FDG imaging in head and neck cancer: a systematic review and meta-analysis. Eur J Nucl Med Mol Imaging 2011;38: Kao J, Vu HL, Genden EM, et al. The diagnostic and prognostic utility of positron emission tomography/computed tomography-based follow-up after radiotherapy for head and neck cancer. Cancer 2009;115: Brenner DJ, Hall EJ. Computed tomography an increasing source of radiation exposure. N Engl J Med 2007;357: Shuryak I, Sachs RK, Brenner DJ. Cancer risks after radiation exposure in middle age. J Natl Cancer Inst 2010;102: Zhang I, Branstetter BF, Beswick DM, Maxwell JH, Gooding WE, Ferris RL. The benefit of early surveillance in HPV-associated head and neck squamous cell carcinoma. Arch Otolaryngol Head Neck Surg 2011;137: Nayak JV, Walvekar RR, Andrade RS, et al. Deferring planned neck dissection following chemoradiation for stage IV head and neck cancer: the utility of PET-CT. Laryngoscope 2007;117: Marur S, D Souza G, Westra WH, Forastiere AA. HPV-associated head and neck cancer: a virus-related cancer epidemic. Lancet Oncol 2010;11: Yao M, Smith RB, Hoffman HT, et al. Clinical significance of postradiotherapy 18 F-Fluorodeoxyglucose positron emission tomography imaging in management of head and neck cancer a long-term outcome report. Int J Radiation Oncology Biol Phys 2009;74: Moeller BJ, Cannon BA, Williams MD, et al. Prospective Risk-Adjusted 18 F-Fluorodeoxyglucose Positron Emission Tomography and Computed Tomography Assessment of Radiation Response in Head and Neck Cancer. J Clin Oncol 2009;27: Ho AS, Tsao GJ, Chen FW, et al. Impact of positron emission tomography/computed tomography surveillance at 12 and 24 months for detection head and neck cancer recurrence. Cancer 2013;119: McDermott M, Hughes M, Rath T et al. Negative Predictive Value of Surveillance PET-CT in Head and Neck Squamous Cell Cancer. AJNR Am J Neuroradiol. 2013;34: