• 9/19/2006
  • Toronto, Ontario, Canada
  • Editorial – John N. Waldron
  • Journal of Clinical Oncology, Vol 24, No 27 (September 20), 2006: pp. 4367-4368

At present, the potential uses for positron emission tomography (PET) in the management of cancer include the characterization of disease at presentation, assessment of disease response to treatment, and the detection of recurrent disease. Published reports of the use of PET in cancer patients listed in the PubMed database have multiplied by a factor of 10 over the last decade, with close to 1,000 reports in 2005.

It is reasonable to conclude that, for the majority of patients at presentation who undergo staging with computed tomography (CT) and/or magnetic resonance imaging (MRI), the use of fluorodeoxyglucose (FDG) PET does not add information that would change initial management. Nonetheless, FDG-PET will detect disease that would have been otherwise overlooked for an important minority of patients. The consequence should prevent undertreatment in the case of specifically targeted therapies such as surgery or radiation. Alternatively, when FDG-PET determines that disease has spread beyond the possibility of curative treatments, the morbidity of overtreatment will be avoided.

A greater potential impact of FDG-PET is its putative ability to exclude disease, thereby avoiding the need for further investigation or more extensive local or systemic treatment. Examples of this potential could be the avoidance of axillary dissection and cytotoxic chemotherapy in breast cancer patients with a PET-negative axilla, the avoidance of mediastinoscopy for lung cancer patients with a PET-negative mediastinum, and, in many cancer sites including the oral cavity, the avoidance of irradiation or dissection of lymph nodes at risk of harboring disease when the PET scan is negative. Evidence of the degree to which FDG-PET may play a role in this regard is just emerging and is appropriately the subject of numerous ongoing clinical trials.

In this issue, Ng et al1 describe the results of a prospective analysis of 134 patients presenting with squamous cell carcinomas of the oral cavity and no evidence of lymph node metastasis on palpation of the neck. All patients underwent FDG-PET imaging along with either MRI or CT imaging before definitive surgical management including neck dissection. The subsequent relationship between the pathologic nodal status and the preoperative imaging is described. Metastatic disease was confirmed histologically in 26% of patients (35 patients, 51 neck levels, and 91 nodes). The sensitivity of FDG-PET compared with CT/MRI was 41% v 22%, respectively, on a level-by-level basis (P = .021) and 51% v 31%, respectively, on a patient-by-patient basis (P = .065). The sensitivity improved marginally to 57% when the FDG-PET was visually correlated with either the MRI or CT images.

If decisions to perform a neck dissection were driven by the neck imaging results alone, 82% of the patients (110 of 134 patients) would have avoided neck dissection. However, for decisions based on CT/MRI, FDG-PET, or visual correlation of both, 21%, 16%, and 14% of patients without neck dissection, respectively, would have been left with disease in the neck. Limiting this analysis to those patients with T1 to T3 tumors (patients at lower risk of subclinical neck metastasis), 10% of patients without neck dissection would have had disease remaining in the neck. It remains a matter for speculation whether these rates would be considered acceptable to most patients and clinicians.

It is apparent that factors other than neck imaging alone must be used in decision making regarding neck dissection in this population. These factors include primary site characteristics such as location, T category, histologic features, and tumor thickness, all of which have been linked to the likelihood of lymph node metastasis. Not surprisingly, Ng et al1 observed that 46% of 24 patients with T4 tumors had nodal metastasis. It is notable that, for more than half of the patients with nodal metastasis, the visual correlation of FDG-PET with either CT or MRI was unable to predict the presence of disease. This resulted in a false-negative rate of 25% for all patients with T4 disease. By selecting the 110 patients with T1 to T3 tumors, a group that had a nodal metastasis rate of 22% (24 of 110 patients), the application of PET with CT/MRI correlation resulted in a false-negative rate of 8%. Although at first glance this seems like an improvement, it must be remembered that the lower false-negative rate is more a reflection of a lower prevalence of neck metastases in this population than an improved ability to detect the metastases.

The degree to which we rely on an investigation with a sensitivity of, at best, 57% (FDG-PET with CT or MRI visual correlation) must be guided by the prevalence and clinical significance of the end point it is intended to detect. With this in mind, the use of an investigation to detect an end point whose prevalence has been reduced through prognostic factor selection to low levels is ultimately self-defeating. One must also keep in mind the consequences of not performing a neck dissection if there is subclinical disease in the regional nodes. Outcome in such instances can be influenced by many factors such as elective neck irradiation, failure at the primary site, the timeliness of the detection of neck failure, and, ultimately, the ability to salvage the patient when nodal metastases are detected. There will be at least a few deaths caused by neck failure that could have been prevented by elective treatment of the clinically negative neck (eg, by radiotherapy or an initial neck dissection). Countering this is the fact that approximately 75% of patients presenting with oral cavity squamous cell carcinoma and clinically N0 necks will have no evidence of metastasis on subsequent histopathologic examination of dissected nodes. Furthermore, the cost of unnecessary neck dissections in both monetary terms and quality of life is not inconsiderable.

The work of Ng et al1 is the largest study of its kind to date, and the 51% sensitivity of FDG-PET alone is among the better results reported in the literature. Before this, most series have been of only modest size (10 to 30 patients) and described sensitivity rates of 30% to 40%.2 Comparison of results between series requires careful consideration of the methods used to define the N0 patient (palpation alone, ultrasound, or CT/MRI) and the methods used to histologically examine the dissected nodes because these will have an impact on reported sensitivities.3 Although most reported series did not use hybrid PET CT scanners (FDG-PET/CT), the incremental gains in sensitivity to be expected using this technology will likely be modest in this patient population. The same issues apply to the present study, which also did not use contemporary hybrid scanner technology.

We are reaching the lower limit of disease detectability with FDG-PET. Although it may seem folly to doubt the seemingly relentless advancement of technology, it does not seem likely that FDG-PET will ever be able to reliably detect the presence of isolated tumor deposits of 3 mm or less. This level of discrimination awaits a future technology. With that in mind, clinicians and patients must carefully consider the results of FDG-PET imaging along with other predictors of the likelihood of occult metastatic spread before making management decisions. For now, FDG-PET or FDG-PET/CT remains but one piece of the puzzle.

References:
1.Ng S-H, Yen T-C, Chang J T-C, et al: Oral cavity squamous cell carcinoma with palpably negative neck: A prospective study of patients evaluated with [18F]fluorodeoxyglucose positron emission tomography and computed tomography and magnetic resonance imaging. J Clin Oncol 24:4371-4376, 2006[Abstract/Free Full Text]

2. Wensing BM, Vogel WV, Marres HA, et al: FDG-PET in the clinically negative neck in oral squamous cell carcinoma. Laryngoscope 116:809-813, 2006[CrossRef][Medline]

3. Brouwer J, de Bree R, Comans EF, et al: Positron emission tomography using [(18)F]fluorodeoxyglucose (FDG-PET) in the clinically negative neck: Is it likely to be superior? Eur Arch Otorhinolaryngol 261:479-483, 2004

Author’s affiliation:
Princess Margaret Hospital, University Health Network, Toronto, Ontario, Canada