Innovations in Radiotherapy Planning of Head and Neck Cancers: Role of PET

Identification of Radiotherapy Targets Based on ¹⁸F-FDG PET

In a recent issue of this journal, Fletcher et al. reviewed the available literature, and a multidisciplinary expert panel developed recommendations on the use of ¹⁸F-FDG PET in oncology practice (3). These recommendations on the use of ¹⁸F-FDG PET for the detection and staging of head and neck tumors are briefly summarized here.

The expert panel concluded that ¹⁸F-FDG PET should not be added to conventional anatomic imaging in the routine diagnostic work-up of primary head and neck tumors (3). This conclusion was drawn because the available data were too uncertain as to whether ¹⁸F-FDG PET can determine the anatomic extent of the primary tumor more accurately than CT or MRI.

Regarding the detection of cervical lymph node metastases, Fletcher et al. concluded that ¹⁸F-FDG PET has a higher sensitivity, specificity, positive predictive value, and negative predictive value than CT and MRI. Therefore, its use in routine local staging was recommended (3). However, this recommendation did not incorporate the findings of the metaanalysis by Kyzas et al. that was published shortly thereafter (4). This metaanalysis reviewed 35 studies using ¹⁸F-FDG PET for the pretreatment evaluation of the lymph node status. The authors concluded that there was no solid evidence to support the routine application of ¹⁸F-FDG PET, as the sensitivity and specificity improved by only 5%−7% compared with conventional imaging modalities. In the subset of studies only enrolling patients without clinically apparent cervical lymph node metastases, the sensitivity was only 50% and not better than conventional imaging methods, specifically ultrasound with fine-needle cytology. From these contradictory recommendations, it is clear that this is an unresolved issue that requires further study.

For the detection of distant metastases, the net benefit of using ¹⁸F-FDG PET was reported to be still uncertain (3). Functional imaging might be beneficial in patients with advanced-stage disease, in whom the odds of having distant metastases are greater (3). In these patients, the ¹⁸F-FDG PET findings may alter the treatment intention from curative to palliative and thus affect the total dose and fractionation scheme. Additionally, it may reduce treatment-related side effects in those patients, as the selected treatment volume is often confined to the primary tumor or metastatic lymph nodes causing discomfort or pain.

The role of ¹⁸F-FDG PET or ¹⁸F-FDG PET/CT for the early detection of recurrent disease is addressed elsewhere (3,5,6).

It can be concluded that there is only a modest role for ¹⁸F-FDG PET in the routine diagnostic workup and staging of patients with head and neck cancer. However, this does not disqualify ¹⁸F-FDG PET as a potentially useful and complementary tool for accurate delineation of the radiotherapy target volume and customized dose delivery.

¹⁸F-FDG PET for Delineation of Radiotherapy Target Volume

In this era of high-precision radiotherapy, accurate tumor volume delineation with respect to tumor boundaries, shape, and volume is crucial. Target volume delineation is primarily based on anatomic information about the tumor and affected lymph nodes. A thorough physical examination of the head and neck forms the basis for assessment of tumor extension, especially for superficially spreading mucosal tumors. Anatomic imaging using CT or MRI provides important complementary information by depicting distorted anatomy and regions of abnormal contrast enhancement. For oral cavity and oropharyngeal carcinomas, MRI is the preferred imaging modality because it achieves better soft-tissue contrast.

There are several potential advantages of using ¹⁸F-FDG PET for target volume delineation. ¹⁸F-FDG PET may reduce the interobserver variability in gross tumor volume (GTV) delineation, reduce the size of the GTV, identify tumor areas or lymph nodes missed by CT or MRI, and identify parts of the GTV potentially requiring an additional radiation dose. However, the use of ¹⁸F-FDG PET also bears some disadvantages: the limited spatial resolution, the lack of a standardized method for signal segmentation, and false-positive ¹⁸F-FDG PET readings caused by inflammation.

A reduction of interobserver variability has been demonstrated for non–small cell lung cancer when ¹⁸F-FDG PET was incorporated in GTV delineation (7,8). In patients with head and neck cancer, this finding has been less consistent. Ciernik et al. investigated the value of ¹⁸F-FDG PET in 39 patients with various solid tumors, of which 12 were head and neck cancer (9). The investigators found both increases (≥25%) and decreases (≤25%) in half the patients when GTV delineation based on CT alone was compared with ¹⁸F-FDG PET/CT. When GTV delineation was compared between 2 experienced radiation oncologists, the mean volume difference of 26.6 cm³ by CT alone was reduced to 9.1 cm³ with ¹⁸F-FDG PET/CT (9). Riegel et al. found conflicting results when 2 experienced radiation oncologists and 2 neuroradiologists delineated 16 patients with head and neck cancer (10). On average, the GTVs based on ¹⁸F-FDG PET/CT were larger than the corresponding CT-based volumes. Furthermore, the authors observed a large discrepancy between the GTV delineation of the 2 radiation oncologists, with one delineating larger volumes on CT and the other on ¹⁸F-FDG PET/CT (10). One important difference between these 2 studies relates to thresholding of the ¹⁸F-FDG PET signal: Ciernik et al. chose a fixed threshold of 50% of the maximum signal intensity, whereas Riegel et al. used a discretionary window-level setting (9,10).

A reduction of the GTV using ¹⁸F-FDG PET has been demonstrated in a landmark study including laryngeal cancer patients (11). The authors investigated the role of coregistered CT, MRI, and ¹⁸F-FDG PET in GTV delineation of patients undergoing laryngectomy. Compared with the reference surgical specimen, ¹⁸F-FDG PET came closest to depicting the true tumor volume. All modalities overestimated the extent of the tumor, ¹⁸F-FDG PET by an average of 29%, CT by 65%, and MRI by 89%. However, all 3 imaging modalities, including ¹⁸F-FDG PET, failed to identify a small fraction of the macroscopic tumor (approximately 10%), mainly consisting of superficial mucosal extensions.

Before PET-based GTVs can reliably and reproducibly be incorporated into high-precision radiotherapy planning, operator-independent segmentation tools have to be developed and validated. Simple visual interpretation of the PET signal is most commonly applied but is highly operator-dependent, as it is susceptible to the window-level settings of the images and interpretation differences (10,12,13). This is why research groups have explored more objective methods, such as isocontouring based on a fixed standardized uptake value (SUV), for example, of 2.5, or thresholds acquired through phantom experiments such as a fixed threshold of the maximum tumor signal intensity (40% or 50%) (9,14,15). Daisne et al. used a variable threshold adaptive to the signal-to-background ratio in their study on laryngeal cancer patients (16). Recently, the same group published a new gradient-based segmentation tool based on watershed transform and hierarchical cluster analysis and validated this in an adaptive biologic image-guided planning study (17,18). Shortly thereafter, van Dalen et al. published an iterative background-subtracted relative-threshold level method validated in patients with liver metastases (19). The optimal relative-threshold level thereby depends on the lesion size but not on the signal-to-background ratio.

A recent study on 78 patients with head and neck cancer compared 5 commonly used methods of ¹⁸F-FDG PET signal segmentation (visual interpretation, 40% and 50% of the maximum tumor signal intensity, fixed SUV of 2.5, and the signal-to-background ratio method; Fig. 2; Table 1) (20). The results showed that the volume and shape of the resulting GTV were heavily influenced by the choice of the segmentation tool. Visual interpretation of the PET signal yielded volumes close to those of CT-based GTV delineation, whereas all automated segmentation methods resulted in significantly smaller GTVs than the GTVs based on clinical information and CT alone (20). Furthermore, in a large percentage of patients (between 29% and 64%, depending on the segmentation tool used) more than 20% of the ¹⁸F-FDG PET–based GTV was located outside the GTV based on clinical information and CT. This suggests that tumor could be identified by ¹⁸F-FDG PET that was missed using the standard methods of GTV delineation. However, in the absence of histologic validation it is unknown in what percentage of cases this was caused by peritumoral inflammation, resulting in a false-positive reading of the ¹⁸F-FDG PET signal.

Most of the discussed data are based on theoretic delineation studies using operator-dependent or -independent segmentation tools. Only the study by Daisne et al. had results validated against histopathology (11). It is obvious that additional validation studies are needed as well as carefully designed clinical trials to address the issue of safety (side effects) and the clinical impact (locoregional control, survival) of incorporating PET for GTV delineation. However, it will be at least challenging to design and conduct such trials while PET/CT is increasingly incorporated into clinical practice on the basis of nonrandomized clinical studies in often relatively small patient populations.

Thus far, most delineation studies incorporating ¹⁸F-FDG PET in head and neck cancer have concentrated on the primary tumor. This is probably due to the fact that CT-based delineation of metastatic lymph nodes is usually less prone to error because of better discrimination from the surrounding fatty tissue. However, this can be more difficult in cases with large, matted nodes. ¹⁸F-FDG PET might be helpful in these situations, although one should be aware of the possibility of negative ¹⁸F-FDG PET readings in necrotic parts of the lymph node.

Apart from more accurate target volume delineation for radiotherapy planning purposes, PET may identify parts of the GTV potentially requiring additional radiation doses. Assuming that ¹⁸F-FDG uptake represents tumor cell density, ¹⁸F-FDG PET can be used to direct dose escalation to ¹⁸F-FDG–avid subvolumes of the tumor. The feasibility of this approach was demonstrated in various theoretic planning studies applying either uniform dose distribution or voxel-intensity–based IMRT (21,22). The former method delivers a uniform escalated dose to an ¹⁸F-FDG–avid subvolume within the CT-based target volume. This approach was pioneered by Schwartz et al., who escalated total dose to 75 Gy in a theoretic planning study involving 20 patients with head and neck cancer (21). With voxel-intensity–based IMRT, the ¹⁸F-FDG signal intensity in the PET voxel is proportionally related to the dose prescribed to that voxel; that is, the higher the PET signal, the higher the prescribed dose (22). Both methods are alternatives for boosting ¹⁸F-FDG PET subvolumes inside a CT-based planning target volume. The clinical feasibility of dose escalation using a uniform dose distribution was recently proven in a phase I clinical trial (23). Forty-one patients with head and neck cancer were treated with IMRT to dose levels of 72.5 and 77.5 Gy using a simultaneous integrated boost. With this technique, the escalated dose is delivered simultaneously with the lower dose to the low-risk areas, as opposed to being delivered sequentially, at the end of the treatment. Acute toxicity (dysphagia ≥ grade 3) occurred in 50% of patients at both dose levels, and dose-limiting toxicity was observed in 2 patients at the lower dose level and one patient at the higher. The authors concluded that PET-guided dose escalation appeared to be well tolerated, with high local control rates in both the lower-dose group and the higher-dose group (85% and 87%, respectively) at 1 y of follow-up (Fig. 3) (23).

During the course of radiotherapy, the tumor volume gradually decreases and one might consider adjusting the GTV and ultimately the radiotherapy dose distribution accordingly. This adjustment could facilitate sparing of normal tissues. For example, during the treatment course for oropharyngeal tumors, the parotid gland is shifted centrally toward the high-dose region by tumor shrinkage and weight loss of the patient (24). As a result, a larger part of the parotid gland is potentially irradiated to a higher dose, possible resulting in a higher incidence and greater severity of xerostomia. Adaptive image-guided radiotherapy using repetitive PET/CT during the course of treatment is a promising approach to adjusting treatment volume and dose distribution, as has recently been demonstrated in a proof-of-principle study (18). Throughout the course of radiotherapy, the GTVs based on ¹⁸F-FDG PET significantly decreased and were at all times smaller than those defined using pretreatment CT and MRI. Radiation treatment planning based on ¹⁸F-FDG PET and volume adaptation progressively reduced the irradiated volumes by 27%−42% (V₉₀–V₁₀₀), compared with traditional CT-based treatment plans obtained before treatment. Disappointingly, this volume reduction only marginally affected the doses to the organs at risk, such as the parotid gland. Adaptive ¹⁸F-FDG PET–guided radiotherapy nevertheless may be an attractive approach, especially for dose escalation strategies.

¹⁸F-FDG PET can provide important complementary information for radiotherapy planning in head and neck cancer. Potentially, the GTV can be reduced on the basis of the PET information, facilitating sparing of nearby normal tissues and allowing dose escalation to relatively small subvolumes. Furthermore, biologic imaging using ¹⁸F-FDG PET may identify areas of tumor spread not recognized by CT or MRI, potentially improving the accuracy of GTV definition. However, to address the clinical value and possible shortcomings of these concepts, additional histologic validation studies and properly designed clinical studies are needed.

Recently, the first results on clinical treatment outcome after integration of ¹⁸F-FDG PET/CT data into IMRT planning have been published. In a case-control study, Rothschild et al. compared 45 patients with stage IV-A pharyngeal carcinomas treated with ¹⁸F-FDG PET/CT–based IMRT with a matched historical cohort receiving standard 3-dimensional conformal radiotherapy (25). The 2-y overall survival and event-free survival rates of patients treated with ¹⁸F-FDG PET/CT–based IMRT were 91% and 80% and significantly better than for the control group. In a similar study, Vernon et al. reported 2-y overall survival and disease-free survival rates of 83% and 71%, respectively, for 42 patients with head and neck cancer of various stages and subsites (26). Toxicity profiles in this second study were reported as favorable.

Even though these initial results are encouraging, they must be interpreted cautiously because they are based on small and heterogeneous patient populations, they have only short follow-up periods, and they use historical controls. Furthermore, it remains unclear from both studies whether the suggested improvements in tumor control can be attributed to improved radiotherapy techniques or are due to the introduction of ¹⁸F-FDG PET/CT or to other factors.

Hypoxia

Hypoxia is a feature of many solid tumors and in particular squamous cell carcinomas of the cervix and the head and neck (27,28). Tumor cell hypoxia can result from 2 mechanisms: limited diffusion capacity of oxygen due to a large distance from the supplying blood vessel (chronic hypoxia), or impaired perfusion of the supplying vessel due to temporary vasoconstriction or endovascular obstruction (acute hypoxia) (29). Treatment modifications are available, but at the cost of increased morbidity (30,31). To individualize treatment and to select patients for these treatment modifications, assessment of the tumor oxygenation status is compulsory. In accessible tumors of the head and neck or uterine cervix, this assessment can be done by invasive polarographic electrode measurements or by immunohistochemical staining of markers in tumor biopsies (32–34). The advantage of the polarographic electrodes is that the entire tumor can be mapped using multiple tracks. However, its clinical use is limited by the invasive nature of the procedure, the restriction to accessible tumors, and the inability to distinguish between normal, necrotic, and tumor tissue. Immunohistochemical staining of tumor biopsy samples results in high-resolution images that can be analyzed for several endogenous and exogenous markers of interest. Unfortunately, the tumor biopsy samples are often small and represent only a fraction of the entire tumor. Furthermore, exogenous markers require intravenous administration before biopsy samples can be taken. Finally, the acquisition of a tumor biopsy often requires the use of general anesthesia, and this procedure is not attractive for repetitive measurement. Noninvasive imaging using PET can provide a spatial map of the intratumoral distribution of hypoxia before and during treatment. This information can potentially be used not only as a selection instrument for treatment modification but also for optimization of radiotherapy planning and delivery.

¹⁸F-fluoromisonidazole (FMISO) is a nitroimidazole PET tracer that is reduced and bound to cell constituents under hypoxic conditions. In the early 1990s, ¹⁸F-FMISO PET was applied in several small clinical trials on different primary tumors (35–37). Since then, ¹⁸F-FMISO PET has been extensively used for the detection of hypoxia in head and neck tumors (38–44). Importantly, in head and neck cancer it was shown that the level of hypoxia depicted by ¹⁸F-FMISO PET before treatment was correlated with locoregional failure (38,42,45). Apart from its prognostic value, Rischin et al. published data supporting the predictive value of ¹⁸F-FMISO PET (42). They performed ¹⁸F-FMISO PET scans in patients with advanced-stage head and neck carcinomas that were treated with radiotherapy and concurrent chemotherapy alone or combined with a hypoxic cytotoxin. Patients with hypoxic primary tumors treated with the additional cytotoxin experienced significantly fewer local failures than patients treated with chemotherapy alone (0 of 8 patients vs. 6 of 9). Furthermore, the absence of hypoxia on ¹⁸F-FMISO PET was associated with a low risk of locoregional failure when treated with chemotherapy alone (42). ¹⁸F-FMISO PET can thus serve as a predictive tool allowing treatment selection based on biologic tumor characteristics. Ultimately, reduction of side effects in patients not benefiting from treatment modification will be feasible.

Apart from tumor characterization, first attempts were made to delineate a biologic target volume and to escalate the dose to the primary tumor based on ¹⁸F-FMISO PET (46–48). Two theoretic planning studies proved the feasibility of dose escalation to the ¹⁸F-FMISO PET–detected hypoxic subvolume using IMRT (46,47). Rajendran et al. demonstrated that, using an IMRT technique, the dose to the ¹⁸F-FMISO PET–detected hypoxic subvolume could be escalated by an additional 10 Gy (47). Lee et al. achieved a dose of 84 Gy in hypoxic areas without exceeding the normal-tissue tolerance (46). Their attempt to further escalate the dose to 105 Gy in hypoxic regions was successful in only 1 of the 2 plans studied. In a third study, Thorwarth et al. compared IMRT planning with dose painting by numbers based on dynamic ¹⁸F-FMISO PET data (48). Thereby, spatially variant doses are delivered to the tumor according to dose-escalation factors determined on the basis of the dynamic ¹⁸F-FMISO PET scan. With this approach, the tumor control probability was increased from 56% to 70% while the same level of toxicity was maintained (48). However, one has to be cautious in interpreting the data because the number of patients included in this study was small.

Until now, clinical experience with hypoxic PET tracers other than ¹⁸F-FMISO has been limited. ⁶⁰Cu(II)-diacetyl-bis(N⁴-methylthiosemicarbazone) was introduced into the clinic after successful preclinical studies demonstrating a strong correlation between tracer uptake and a low level for partial pressure of oxygen (49). It was the first hypoxia-related PET tracer for which the potential use of a selective boost to the hypoxic subvolume was illustrated (50). However, partly because of its limited specificity, especially if imaging is performed at early time points after administration, this compound did not find its way into larger-scale clinical studies.

¹⁸F-fluoroerythronitroimidazole, ¹⁸F-fluoroazomycin arabinoside, and ¹⁸F-2-(2-nitroimidazol-1-yl)-N-(3,3,3-trifluoropropyl)-acetamide (¹⁸F-EF3) are members of a new generation of nitroimidazoles. ¹⁸F-fluoroerythronitroimidazole showed a higher and more heterogeneously distributed tracer uptake in tumors than in adjacent neck muscle (51). Furthermore, a high uptake of ¹⁸F-fluoroerythronitroimidazole before radiation therapy was associated with a trend toward poor overall survival (52). ¹⁸F-fluoroazomycin arabinoside has similar tracer characteristics to ¹⁸F-fluoroerythronitroimidazole and was proven feasible and of sufficient quality for clinical use in patients with head and neck cancer (53,54). Grosu et al. incorporated ¹⁸F-fluoroazomycin arabinoside PET into radiation treatment planning and detected hypoxic subvolumes of different sizes and distributions (representing on average 11% of the primary tumor volume and 8% of the metastatic lymph node volume) (55). Dose escalation to 80.5 Gy in ¹⁸F-fluoroazomycin arabinoside PET–detected hypoxic areas was shown to be feasible. ¹⁸F-EF3 was used in a phase I study of patients with head and neck cancer (56). In that study, the use of this tracer was shown to be safe, but the number of advanced-stage tumors showing increased tracer uptake was disappointingly low.

In summary, although numerous hypoxic or hypoxia-related PET tracers are available for clinical use, their prognostic and predictive value needs to be assessed in larger clinical studies before implementation for patient selection. Preferably, the PET tracer used must also visualize changes in the oxygenation status caused by treatment modifications counteracting hypoxia, such as carbogen breathing (57). More important, the concept of dose painting to hypoxic subvolumes either by uniform doses or by dose painting by numbers is still the subject of intense debate. There are major concerns about the spatial resolution of hypoxic PET when compared with the distribution and fluctuation of tumor cell hypoxia at the microregional level. In this context, we investigated 10 different head and neck carcinoma xenograft tumor lines using ¹⁸F-FMISO autoradiography and pimonidazole immunohistochemistry (Fig. 4) (58). We found that the pattern of the ¹⁸F-FMISO signal depended on the distribution of hypoxia at the microregional level. In 5 xenograft tumor lines, a significant correlation between the mean ¹⁸F-FMISO and pimonidazole signal intensities was found, and this correlation depended on the underlying microarchitecture. This finding indicates that one should be cautious when studying small tumor subvolumes for dose escalation (58). Apart from different distribution patterns of hypoxia at the microregional level, one has to consider that the oxygenation status changes during the course of radiotherapy, making repetitive PET imaging before and during treatment compulsory (59). Finally, the question on the radiation dose levels required for effective elimination of the radioresistant subpopulations remains unanswered.

Tumor Cell Proliferation

The major limitations of ¹⁸F-FDG PET in oncology are false-positive readings due to tracer uptake in inflammatory tissue or reactive lymph nodes. Therefore, PET tracers that more specifically image DNA synthesis are being developed and tested.

Tumor cell proliferation during the course of therapy adversely affects radiation treatment outcome and prognosis in squamous cell carcinomas of the head and neck (60). 3′-deoxy-3′-¹⁸F-fluorothymidine (¹⁸F-FLT) is a tracer that reflects the activity of thymidine kinase 1, a principal enzyme in the salvage pathway of DNA synthesis (61). The ¹⁸F-FLT PET signal is more specific for actively dividing tumor cells than is the ¹⁸F-FDG PET signal. Inflammatory cells near the tumor consume glucose and thus cause false-positive ¹⁸F-FDG PET readings. However, as these immune response cells are terminally differentiated, the DNA synthesis rate and therefore the ¹⁸F-FLT uptake are not increased. ¹⁸F-FLT PET was validated against histopathology in a variety of solid tumors including breast, lung, and sarcoma (62–64). In soft-tissue sarcoma, Cobben et al. found a significant correlation between the SUVs and labeling index of the proliferation marker Ki-67. In addition, ¹⁸F-FLT PET was able to distinguish low-grade from high-grade soft-tissue sarcomas (62). In breast tumors, Kenny et al. reported a strong correlation between SUVs and the fully quantitative net irreversible plasma to tumor transfer constant (K_i) parameter of dynamic ¹⁸F-FLT PET and staining of Ki-67 (63). Finally, Yap et al. also observed a significant correlation between ¹⁸F-FLT uptake in non–small cell lung cancer lesions and the Ki-67 labeling index (64). In primary head and neck tumors, this promising compound has thus far been applied only to primary laryngeal tumors (65). Validation of ¹⁸F-FLT PET in a large series of squamous cell carcinomas of the head and neck is ongoing at our center. For the detection of cervical lymph node metastases, first results demonstrated that ¹⁸F-FLT PET is not suitable in this tumor entity (66). A high rate of false-positive findings caused by ¹⁸F-FLT uptake in the germinal centers of reactive lymph nodes resulted in a low specificity and a low positive predictive value (17% and 38%, respectively) (66).

Until the present, adaptive image-guided radiotherapy has been based on repetitive PET scanning using ¹⁸F-FDG (18). As the treatment course progresses, the obtained ¹⁸F-FDG PET signal is heavily influenced by the inflammatory response of tumor-surrounding tissues, leading to an increased background activity. As a result, segmentation of the PET signal for tumor delineation purposes becomes increasingly difficult. The use of a proliferation-specific PET tracer, such as ¹⁸F-FLT, may be a solution to this problem. During the course of therapy, the reduction in the proliferative activity of the primary tumor can be accurately imaged by ¹⁸F-FLT, not disturbed by increased tracer uptake in surrounding inflammatory tissue. Our group is assessing the changes in the ¹⁸F-FLT PET signal during therapy in patients with squamous cell carcinomas of the head and neck treated with radiotherapy alone or with concomitant chemotherapy (Fig. 5). The predictive potential of this approach and applicability for tailored treatment are the subject of investigation.

Perfusion, Protein Synthesis, and Others

Another significant tumor characteristic strongly related to tumor cell hypoxia is tumor blood perfusion. Hypoxia is a strong stimulus for neovascularization, but many newly formed vessels are of poor quality and have severe structural and functional abnormalities. Despite increased vascular density, the impaired functionality of blood vessels may result in deprivation of oxygen and nutrients. Therefore, an imaging tool for assessment of tumor blood flow may provide important information relevant for radiotherapy responsiveness. Lehtiö et al. used ¹⁵O-labeled water (¹⁵O-H₂O) and ¹⁸F-fluoroerythronitroimidazole for imaging of perfusion and hypoxia in 21 patients with head and neck cancer (51). Preliminary results from this small study indicated an association between tumor perfusion and radiation treatment outcome (52).

O-2-¹⁸F-fluoroethyl-l-tyrosine (¹⁸F-FET) and l-methyl-¹¹C-methionine (¹¹C-MET) are amino acid analogs used to visualize cellular amino acid uptake or protein synthesis. ¹⁸F-FET may be useful in differentiating tumor from posttreatment inflammatory tissue, as it is not taken up by inflammatory cells. Several studies compared ¹⁸F-FET with ¹⁸F-FDG PET in squamous cell carcinomas of the head and neck and histopathologically confirmed the specific uptake of ¹⁸F-FET by malignant cells (67–69). The specificity of ¹⁸F-FET PET was found to be superior to that of ¹⁸F-FDG PET (95%−100% vs. 63%−79%), but the sensitivity of the amino acid tracer was significantly lower (64%−75% vs. 93%−95%, respectively). Because the SUVs for ¹⁸F-FET PET were significantly lower than those for ¹⁸F-FDG PET, the new tracer will probably not replace ¹⁸F-FDG PET as a diagnostic tool but can provide complementary information for discrimination between tumor and inflammatory tissue (67–69).

¹¹C-MET PET is similar in sensitivity and specificity to ¹⁸F-FDG PET (70). In a delineation study, ¹¹C-MET was compared with ¹⁸F-FDG PET and CT. Although, compared with CT, ¹⁸F-FDG PET yielded significantly smaller GTVs, GTVs based on ¹¹C-MET PET were not different from GTVs based on CT, probably because of uptake by surrounding normal mucosa and salivary gland tissue. The authors concluded that ¹¹C-MET has no additional value for target volume delineation in head and neck tumors (71).

1-¹¹C-acetate is suggested to preferentially metabolize to the membrane lipids in tumor cells. In a staging and radiotherapy planning study for head and neck cancer, 1-¹¹C-acetate PET detected all primary tumors and 95% of the metastatic lymph nodes, more than ¹⁸F-FDG PET and CT/MRI (72). However, the GTVs derived by 1-¹¹C-acetate PET were 51% larger than those based on ¹⁸F-FDG PET. Before 1-¹¹C-acetate PET can be introduced in the radiotherapy planning process, further studies are needed to explain this discrepancy and to clarify the mechanism of tumor uptake.

Finally, noninvasive methods to assess the uptake and biodistribution of biologic modifiers will be of great value to direct new targeted therapies. Radiolabeled antibodies and small molecules for PET are currently being developed and tested in preclinical and early clinical studies (73,74).

In conclusion, PET tracers that image specific biologic tumor characteristics offer potential for tailor-made radiation therapy. However, they remain in the research arena until proper clinical validation has occurred.