Can Evaluation of Targeted Therapy in Oncology Be Improved by Means of ¹⁸F-FLT?

One of the most promising aspects of molecular imaging is its potential capacity to measure therapy effects long before morphologic changes are detected. The most frequently used PET tracer in oncology remains ¹⁸F-FDG. However, despite its high sensitivity, this tracer has some major drawbacks, of which the generally low specificity is the main important limitation. Therefore, other, more specific tracers have been evaluated. One of the most promising and thoroughly studied radiopharmaceuticals is the proliferation marker 3′-deoxy-3′-¹⁸F-fluorothymidine (¹⁸F-FLT). Accumulation of ¹⁸F-FLT in tumor cells has See page 1559been shown to be dependent on cellular thymidine kinase-1 activity, the key enzyme and the limiting step of the pyrimidine salvage pathway of DNA synthesis, which is overexpressed in most tumor types (1). After monophosphorylation of ¹⁸F-FLT by thymidine kinase-1, ¹⁸F-FLT is intracellularly trapped. Because thymidine kinase-1 is functional only in the late G1- and S-phase of the cell cycle, ¹⁸F-FLT uptake closely correlates to the amount of proliferating cells (2,3).

Although ¹⁸F-FLT uptake is generally lower than ¹⁸F-FDG uptake, making it unlikely that ¹⁸F-FLT will replace ¹⁸F-FDG for staging purposes, the higher specificity of this tracer and lower false-positive rate is a major advantage for tumor grading and early response assessment. The main cause of the limited specificity of ¹⁸F-FDG is the high uptake in inflammatory cells, which cannot be differentiated from malignant cells. A much lower uptake in inflammatory tissue was shown for ¹⁸F-FLT than for ¹⁸F-FDG (4). However, the initial enthusiasm about the higher specificity of ¹⁸F-FLT has been tempered by recent reports that ¹⁸F-FLT uptake also occurs in granulomatous inflammatory lesions such as tuberculosis (5) and in reactive lymph nodes (6), being related to a high proliferation rate of macrophages and B-lymphocytes, respectively. Next to the inflammatory processes, which can be mistaken for tumors, transient inflammatory changes can occur inside a tumor as a reaction to therapy, inducing a temporarily increased ¹⁸F-FDG uptake. Thereby, measurement of ¹⁸F-FDG uptake could result in an underestimation of therapy response. Because inflammatory cells have a much lower proliferation rate, proliferation tracers such as ¹⁸F-FLT will be less hampered by this phenomenon, and measurements of tracer uptake will more accurately reflect tumor response. A recently published study confirmed that a temporary rise in inflammatory cells after cyclophosphamide administration did not significantly influence ¹⁸F-FLT uptake, whereas ¹⁸F-FDG uptake was temporary increased (7).

Besides the issue of inflammatory response, many new cancer-treatment agents induce cell-cycle arrest instead of tumor cell death and are not expected to lead to fast tumor regression. This issue makes measurements of cellular viability by ¹⁸F-FDG theoretically less relevant, and the assessment of cellular proliferation by ¹⁸F-FLT might be a valid alternative. Disease-specific molecularly targeted agents increasingly replace the empiric combinations of cytotoxic agents from the past, because cytotoxic agents frequently lead to resistance and with each subsequent relapse the response rate will decrease (8). An example of the current targeted strategies is inhibition of the mammalian target of rapamycin (mTOR). Several analogs, such as temsirolimus (CCI-779; Wyeth), everolimus (RAD-001; Novartis), and deforolimus (AP23573;ArIAD and Merck), are being tested in clinical trials for treatment of mantle cell lymphoma, ovarian cancer, neuroendocrine carcinoma, and endometrial carcinoma (8,9). mTOR is a regulator of cellular proliferation and acts through several targets. One of these targets is the messenger RNA encoding for the cyclin D1 protein, involved in cell-cycle regulation. Blocking mTOR leads to an inhibition of the translation of cyclin D1 messenger RNA to the cyclin D1 protein and provokes cell-cycle arrest in mid to late G1 (10), before the upregulation of thymidine kinase-1 in the S-phase, and thereby directly influencing ¹⁸F-FLT uptake in the cell. As a result, imaging of proliferation with ¹⁸F-FLT can directly measure the effect of mTOR inhibition and distinguish patients responding to mTOR inhibition from patients experiencing only the side effects of the therapy.

In half of all advanced ovarian cancers, p53 is mutated. This mutation is associated with a lack of response to cisplatin therapy, and as a result many of these patients have incurable disease (11). Previous studies showed that inhibition of mTOR blocks ovarian cancer cell proliferation and enhances the effect of cisplatin (12). ¹⁸F-FDG has been shown to be ineffective in predicting response to mTOR inhibition in patients with solid tumors (13). In this issue of The Journal of Nuclear Medicine, Aide et al. present an important preclinical study on ¹⁸F-FLT PET after mTOR inhibition in a cisplatin-resistant ovarian tumor model (14). Aide et al. aimed to evaluate ¹⁸F-FLT PET during a daily administered everolimus therapy. ¹⁸F-FLT uptake was correlated to bromodeoxyuridine uptake as a marker of cell proliferation and phosphorylation of ribosomal protein S6 as a downstream marker of mTOR activation. ¹⁸F-FLT uptake decreased 2 d after initiation of treatment, with a more pronounced effect at day 7 of mTOR inhibition. Correlative immunohistochemistry showed a marked decrease in pS6 activity and bromodeoxyuridine incorporation corresponding to the decreased level of ¹⁸F-FLT uptake. In this preclinical feasibility study no correlation with outcome, or in other words the value of ¹⁸F-FLT for the prediction of response, was studied. However, this aspect of molecular imaging, especially, should be further evaluated to really demonstrate its usefulness in daily routine.

A recently published study described ¹⁸F-FDG and ¹⁸F-FLT imaging after a single dose of temsirolimus or cyclophosphamide in a mouse model of mantle cell lymphoma (7). ¹⁸F-FLT uptake decreased early after mTOR inhibition, in correlation with cyclin D1 expression, which dropped from day 1 until day 4. However, on day 7 after mTOR inhibition a temporary rise was observed in ¹⁸F-FLT uptake and cyclin D1. It is possible that still-viable tumor cells reenter the S phase after removal of the drug (half-life of temsirolimus is 9–17 h). Additionally, it is important to notice that ¹⁸F-FLT is only slightly incorporated into DNA and that thymidine kinase-1 may be upregulated despite an inhibition of the DNA synthesis. ¹⁸F-FLT uptake might also be stimulated by cellular repair mechanisms or the salvage pathway of the pyrimidine metabolism (15). An in vitro study observed an early increase in ¹⁸F-FLT uptake 24 h after 5-fluorouracil due to blocking of the de novo pathway of the pyrimidine metabolism, thereby inducing the salvage pathway and redistributing nucleoside transporters to the plasma membrane (16). In the study of Aide et al. (14), no temporary ¹⁸F-FLT rise was observed during therapy. This observation is likely due to the permanent mTOR inhibition, as everolimus was administered daily in this study whereas the ¹⁸F-FLT rise was observed 7 d after a single treatment.

Although a temporary increase in ¹⁸F-FLT signal after treatment should be taken into consideration, it is clear that the paper from Aide et al. supports the concept of early response assessment with ¹⁸F-FLT. In this context, a recent paper reported an ¹⁸F-FLT decrease after cytotoxic chemotherapy in patients with metastatic germ cell tumor (17). However, no significant differences were observed between histologic responders and nonresponders. In conclusion, the current evidence suggests that ¹⁸F-FLT monitoring is more likely to be successful in patients undergoing cytostatic therapy. Of course, this possibility will have to be evaluated in larger clinical trials for different tumor entities.

• © 2010 by Society of Nuclear Medicine

• Received for publication June 4, 2010.
• Accepted for publication July 2, 2010.