What Is the Role of PET-CT in Residual Occult Nodal Disease?

PHOENIX-Just how useful are PET or PET-CT in identifying residual occult nodal disease in patients with head and neck cancer? Two presentations at the 2009 Combined Otolaryngology Spring Meeting addressed this topic. In one, researchers suggested that delaying PET-CT may work to help reduce the number of planned neck dissections. In the other, researchers said that PET-CT does have a role to play in head and neck patients, but cannot replace neck dissection as a staging tool.

Christine Gourin, MD, Associate Professor of Otolaryngology-Head and Neck Surgery at Johns Hopkins University in Baltimore, presented details of a retrospective study of patients treated for head and neck squamous cell cancer (HNSCC).

The role of PET-CT in detecting residual occult nodal disease in patients following chemoradiation is controversial, she said. Generally, what is advocated is a planned post-treatment neck dissection-regardless of clinical response to chemoradiation. Proponents of neck dissection argue that there is a high incidence of residual occult disease found in histopathological evaluations of samples from neck dissection, and that the procedure leads to better regional control rates. Opponents, however, maintain that there are low regional recurrence rates in patients with a complete response who are not treated with neck dissection, and that there is no change in survival in most patients by adding neck dissection.

The question is, can PET-CT help identify patients who could be spared unnecessary neck dissection? The combined technology provides both physiologic and anatomic information. But the efficacy of PET-CT in predicting residual nodal disease after chemoradiation, and thus the need for posttreatment neck dissection, appears dependent on the timing of imaging after chemoradiation, Dr. Gourin said. Post-treatment neck dissection is ideally performed within four to 12 weeks after chemoradiation. This study sought to determine if PET-CT findings correlated with neck dissection findings.

PET-CT Study

In the study she presented, researchers reviewed the records of all patients with advanced (N2 or N3 disease) HNSCC who had been treated with chemoradiation from December 2003 to June 2007. Patients who had a complete response to treatment and underwent both PET-CT and a planned post-treatment neck dissection comprised the study group. A total of 32 patients were included, 28 males and four females, with a mean age of 56 years. All patients had stage IV disease at presentation. Mean follow-up after completion of chemoradiation was 20 months.

None of the patients had clinical signs of neck disease when PET-CT or neck dissections were performed. PET-CT showed positive results for residual nodal disease in 20 patients. These were compared to pathological findings from samples taken from neck dissection. Residual carcinoma was found in six of the 20 patients (30%) who had positive PET-CT findings, and in four of 12 patients (33%) who had negative PET-CT findings.

Among the 14 patients who had positive PET-CT findings but no tumor, it was found that five patients had extensive necrosis, and either a histiocytic response or foreign body giant cell reaction was found in nine patients. Of the eight patients in whom PET-CT and pathology studies were both negative for tumors, three patients were found to have radiation fibrosis, and five had a foreign body reaction with giant cells and cholesterol debris.

Overall, the sensitivity and specificity of PET-CT in predicting occult nodal disease was 60% and 36% respectively, Dr. Gourin said, with a positive predictive value of 30%, and a negative predictive value of 67%. PET-CT performed within three months of treatment did not correlate with pathologic findings.

Researchers also took into account the standard uptake value (SUV) of PET and how this may have related to findings. SUV levels show how active cells are in terms of utilizing the FDG contrast material; however, both nonviable tumor cells and areas where inflammation is occurring show increased SUV, which can increase the false positives from PET imaging when PET-CT is performed too early, she said.

Timing of PET-CT Also a Factor

Whether post-treatment neck dissection should be used in HNSCC with advanced disease is controversial because not all patients have residual occult disease. But if surgery is to be performed after chemoradiation, evidence suggests that the safest window is four to 12 weeks after chemoradiation. Unfortunately, this window corresponds to a time when PET-CT has a lower accuracy-studies show that performing the imaging later, at about 12 weeks after inflammation has been reduced and there may be an increase in tumor size, is associated with greater accuracy, she said.

Deferring early post-treatment neck dissection in patients without residual adenopathy in favor of obtaining PET-CT imaging in 12 weeks, or later, following chemoradiation may be a viable approach to help reduce the number of unnecessary neck dissections, Dr. Gourin said.

The regional recurrence rate after planned neck dissection was 6%, which is similar to rates reported for patients with a complete response who are observed with serial negative PET-CT imaging. Overall, our data suggest that not all residual viable tumors in post-treatment neck dissections are viable and that the timing of PET-CT should influence the decision to proceed with neck dissection, not the other way around, she said.

Another Opinion

Another take on the subject came from Cherie Ryoo, MD, a fourth-year resident in otolaryngology-head and neck surgery at Ohio State University, who presented findings from a retrospective study of 243 head and neck cancer patients who had a diagnosis of upper aerodigestive tract squamous cell cancer. Data were collected for patients who were treated from January 2005 to December 2007.

Criteria for patients included a neck dissection as part of their treatment, and a whole-body PET-CT performed prior to treatment. The researchers used an SUV value of 2.5 as a standard for positive findings, and compared this to data on histopathological findings.

The researchers found that in the end, in the group of subjects with N0 necks, there were 26 true positives and 80 true negatives. If we based our decision to perform a neck dissection solely off the findings of the PET-CT, we would have performed 18 unnecessary dissections and missed 20 positive specimens. For comparison, there were the patients that had an end-positive neck-these would have resulted in six unnecessary dissections and 10 missed positive specimens, she said. PET-CTs were done prior to treatment.

Dr. Ryoo concluded that PET-CT does not currently have the capability of replacing neck dissection as a staging tool. The technology has room for improvement.

©2009 The Triological Society