18F-5-FPN: A Specic Probe for Monitoring Photothermal Therapy Response in Malignant Melanoma

The increasing global burden and the signicant breakthroughs in malignant melanoma therapy make urgent demands on ecient response evaluation and surveillance of adverse events. Though there have been a few probes explored for early diagnosis or staging of malignant melanoma, but rare for response assessment investigations except for common 18 F-deoxyglucose ( 18 F-FDG). Thus, this research would further explore the feasibility and ability of 18 F-5-uoro-N-(2-(diethylamino)ethyl)picolinamide ( 18 F-5-FPN)PET imaging to evaluate photothermal therapy (PTT) response of malignant melanoma, simultaneously comparing with 18 F-FDG. X.L.L.; Data curation, Y.C.W., M.T.L., W.X.W., H.J., C.Y.J., N.H. and H.L.L.; Formal analysis, Y.C.W. and M.T.L.; Methodology, Y.C.W., M.T.L. and X.D.X.; Software, Y.C.W., M.T.L., W.X.W., H.J. and C.Y.J.; Supervision, Y.X.Z. and X.L.L.; Validation, Y.C.W., M.T.L., W.X.W., H.J., C.Y.J., N.H. and H.L.L.; Writing—original draft, Y.C.W. and M.T.L.; Writing—review and editing, Y.C.W., M.T.L., W.X.W., H.J., C.Y.J., N.H., H.L.L., X.D.X., H.L.J., Y.X.Z. and X.L.L.; Project Administration, Y.X.Z. X.L.L.; Funding Acquisition, X.L.L. read nal

potential probe may provide a new approach for precise and useful response evaluation, timely therapeutic regimen management, and sensitive follow-up.

Background
Melanoma is an aggressive malignant skin neoplasm. Its global incidence is increasing, and Australasia, North America, and Europe are the regions with the highest incidence and mortality [1]. Currently, drug therapy of malignant melanoma (MM) has been expanded from conventional chemotherapy to targeted therapy and immunotherapy [2,3]. Compared with the limited bene ts of systemic chemotherapy, targeted therapy has a notable e cacy on the majority of patients with BRAF mutation. Immunotherapy resulted in better response than chemotherapy (26-40% vs. 4-14%) in a phase III randomized controlled trial [3][4][5]. However, drug resistance occurs in targeted therapy, and immunotherapy works slowly and triggers immune-related adverse events (rash, diarrhea, colitis, etc.) [4,5]. The increasing global burden and intricate situation of novel therapies of MM make more urgent and stricter demands on e cient response evaluation.
Computed tomography (CT) and magnetic resonance imaging (MRI) are the most common techniques for therapy assessment in the clinic and the shrinkage of tumor size suggests a response. A typical response may present as an initial increase in size followed by regression later, i.e., "pseudoprogression" [6,7]. Conventional imaging techniques have di culty in distinguishing real progression from pseudoprogression. Immunotherapy usually takes a long time to observe a response, so that conventional imaging techniques cannot identify early responses, and repeated inspections are inevitable, prolonging the cycle of response evaluation [6,8]. It is well-known that changes in function and metabolism always precede changes in tumor size. Functional imaging [such as positron-emission tomography (PET) imaging] is able to detect the early changes in the function and metabolism of tumors during treatment [7]. 18 F-uorodeoxyglucose ( 18 F-FDG) PET/CT imaging has a vital role in evaluating lesions' response to conventional chemotherapy but has limits for immunotherapy because the high glucose consumption of immune response may obscure the tumor response [8]. Both MM and in ammation accumulate 18 F-FDG, making it di cult to discriminate between pseudoprogression and real progression. One study utilized 18 F-FDG PET imaging to monitor the response of MM patients to anti-PD-1 therapy, nding that three out of eight cases of high uptakes were biopsy-con rmed in ammatory in ltration, and ve were MM [9].
Therefore, more speci c radiopharmaceuticals are needed for response evaluation [8].
PTT is a therapy that utilizes heat to eliminate tumor cells with laser light via photothermal conversion [15]. Melanin, usually abundant in MM, has a broad absorption spectrum (200-1200 nm) and transforms the absorbed optical energy into heat almost entirely [16,17]. Stritzker et al. delineated that the temperature of an amelanotic tumor cell suspension enhanced only 3 ℃ when 808 nm laser irradiated for 2 min, while the temperature of pigmented cells elevated 41 ℃ under the same conditions [18]. The near-infrared I window (650-950 nm) is absorbed least in normal tissues and penetrates tissue deeply [15]. Therefore, we used an 808 nm laser for PTT to treat MM in mice, followed by subsequent response surveillance studies via 18 F-5-FPN and 18 F-FDG PET imaging.
Design of animal models BALB/c nude mice (female, 4 weeks, speci c pathogen-free, HFK Bioscience Co., Ltd.™, Beijing, China) were raised under speci c pathogen-free facilities with a constant temperature and humidity, and free availability of food and water at 12 h/12 h light/dark cycle, approved by the Laboratory Animal Care of Huazhong University of Science and Technology, and in compliance with the regulations and standards of the Institutional Animal Care and Use Committee of Tongji Medical College of Huazhong University of Science and Technology. All mice adjusted to the new conditions for at least 3 d before an arrangement.
B16F10 (1 × 10 6 ) and MDA-MB-231 (5 × 10 6 ) cells in 75 µL phosphate buffered saline were subcutaneously injected into the right forelimb of nude mice. B16F10 models for histological analysis were prepared by injecting cells in both forelimbs with one treated and the other untreated. Subsequent studies were carried out when the greatest diameter of the tumors reached approximately 5 mm. served as an in ammatory model. It was followed by 18 F-FDG and 18 F-5-FPN small animal PET scanning at 24 h and 54 h after injection, respectively [19,20].

Photothermal therapy (PTT)
B16F10-and MDA-MB-231-xenografted mice were randomized to treated or untreated groups. The treated groups were anesthetized intraperitoneally with 2% pentobarbital sodium solution (80 mg/kg), and placed in the left lateral decubitus position to expose the tumors to the 808 nm laser for 10 min (1 W at a distance of 1.5 cm). During the entire procedure, an IR thermal imaging camera (EasIR-4; Gaode Infrared Limited Company, Wuhan, China) was used to monitor the temperature changes in the tumors continuously. The maximum temperatures were recorded at 20 s intervals in the initial 3 min, followed by 1 min intervals. The day of therapy was termed as Day 0.

PET imaging and analysis
Mice fasted for 6-8 h before 18   Imaging surveillance of B16F10 models 18 F-FDG PET scanning on Day − 2, 1, 5, 8, 16 and 18 F-5-FPN PET imaging on Day − 1, 2, 6, 9, 17 were performed on B16F10-xenografted treated and untreated groups (n = 7 per group). Mice were photographed at all imaging time points. If a mouse died during the surveillance or after the last scan, it was dissected to assess whether lung and liver metastases occurred or not, with lung, liver and kidney analyzed via H&E staining.
Survival analysis B16F10-and MDA-MB-231-xenografted mice were randomized to treatment or no treatment (n = 10 per group). The treated group was administered PTT protocol. The condition and weight of mice and the tumor size were monitored every other day, and tumor volume was calculated with the formula (length × width 2 × π/6). If tumor volume was > 2000 mm 3 , or if the loss of weight was > 20% of baseline, or if disability of movement, eating, or drinking was observed, mice were euthanized by double 2% pentobarbital sodium solution. Then, the mice were dissected to evaluate whether lung and liver metastases occurred or not.
Histological analysis B16F10-xenografted tumors in the treated and untreated groups, MDA-MB-231-xenografted tumors, in amed muscles, normal muscles, lungs, livers, and kidneys were xed in 4% paraformaldehyde, embedded in para n, and cut into 4 µm slices for H&E staining.

Statistical analysis
All quantitative data are expressed as mean ± standard deviation. Two group comparisons were conducted by paired or independent t-tests and multiple group comparisons by one-way analysis of variance (IBM SPSS Statistic 20) with p value < 0.05 representing signi cance. Kaplan-Meier plots were used to exhibit survival data, and a log-rank test was used to analyze them (GraphPad Prism 6.0 GraphPad Software™, La Jolla, CA, USA).

Photothermal therapy (PTT)
B16F10-xenografted and MDA-MB-231-xenografted mice were irradiated with the 808 nm laser. Over 13 min, an IR thermal imaging camera was used to monitor and record the temperature continuously. The temperature of the B16F10 tumors increased rapidly with the increment of 30 °C in 1 min and nally exceeded 60 °C, while the MDA-MB-231 tumors heated up minimally (< 5 °C) (Fig. 1a, b). The near-infrared images of B16F10 tumor before PTT, and of B16F10 and MDA-MB-231 tumors during PTT, are presented in Fig. 1c. Immediately after PTT, B16F10 tumors appeared smaller, accompanied by a mild circular burn of the surrounding skin due to thermal conduction, while no similar change was found in MDA-MB-231 tumors (Fig. 1d).

Histological analysis of B16F10 tumors
The BALB/c nude mice (n = 5) injected with B16F10 cells in both forelimbs underwent PTT on one side.
One day after PTT, both tumors in the treated and untreated sides were collected for H&E staining. Tumors in the untreated group (Fig. 1f) possessed an intact structure and sharp borders. Tumor cells were distributed compactly with visible cellular morphology and brown pigment inside. Conversely, tumors in the treated group (Fig. 1e) showed disordered structure, a large amount of necrosis and pyknosis, and no clear borders. A distinct boundary between the treated and surrounding normal skin on the treated side is shown in Fig. 1g.
Imaging surveillance of B16F10 models The treated and untreated B16F10-xenografted mice (n = 7 per group) underwent 18 F-FDG and 18 F-5-FPN PET scans at different set time points (Fig. 2a). Figure 2b depicts the number of surviving mice at each time point. 18 F-5-FPN and 18 F-FDG images were capable of monitoring the response to PTT in the treated group at different time points (Fig. 2c). Tracer accumulation of tumor sites in the early stage after treatment (Day 1, 2) was lower than before treatment (Day − 2, -1). Although no tumor recurrence was observed in the treated group on Day 5 or 6, PET images showed tracer accumulation at the primary site.
On the subsequent scans, tumor relapse was found, with high uptakes. Tumors in the untreated group grew quickly and developed local necrosis, and corresponding 18 F-5-FPN and 18 F-FDG PET images showed increasing tumor size and initial high uptake followed by an absence of uptake in the necrotic region (Fig. 2d).
Quantitative comparisons among the different groups at different time points are shown in Fig. 2e-h and Table 1. 1) The mean tumor uptakes of 18 F-5-FPN in the treated and untreated groups were very high before treatment. In the treated group, uptake dropped initially after treatment, and then slowly increased, while the values in the untreated groups kept on high level rst and diminished at the last scan (Fig. 2e).
The uptake values of 18 F-FDG in the treated and untreated groups were relatively low on the day before treatment, gradually increasing over time in the treated group, while the degree of uptake in the untreated group peaked early and then gradually decreased (Fig. 2f). 2) Comparison between treated and untreated groups on the same day (Fig. 2e, f and Table 1) showed the mean tumor uptakes of 18 F-5-FPN and 18 F-FDG were similar before treatment (p > 0.05). After treatment, all values differed between the two groups.
3) Values of the tracer uptake before treatment and at each time point after PTT in the treated group are also shown in Fig. 2g, Fig. 2h and Table 1. The mean tumor uptakes of 18 F-5-FPN on Day 2 and Day 6 after PTT were much lower than that value before treatment (P < 0.01). However, no signi cant difference existed between the 18 F-FDG uptakes on Day 1 after PTT and before treatment (P > 0.05). Then the tumor recurred and both tracer accumulations increased. Mice were dissected if they died during the surveillance or were euthanized and then dissected after completion of imaging. Tissue samples of lung, liver and kidney were xed with 4% paraformaldehyde, embedded, and cut into slices for H&E staining. No lung or liver metastases or renal dysfunction were found (Fig. 3).

Survival analysis
In the survival analysis, B16F10-xenografted and MDA-MB-231-xenografted mice had been randomized to treated and untreated groups. Six days after PTT, B16F10 tumors (n = 10) in the treated group had shrunk dramatically and formed eschar. Then, four mice had no tumor relapsed and only local pigmented dots remained (Fig. 5a "Suppressed" and g "B16F10-Suppressed"). However, tumors of six other treated mice recurred and developed rapidly in a way similar to that in the untreated group (Fig. 5a "Relapse" and g "B16F10-Relapse"). As expected, B16F10 tumors (n = 10) in the untreated group grew rapidly (Fig. 5a "Untreated" and g "B16F10-Untreated"). MDA-MB-231 cells had been subcutaneously injected into twenty nude mice and nineteen tumor models succeeded with one failing. One mouse was excluded due to a large hematoma that interfered with tumor measurement, and so 18 mice were randomized to receive treatment or no treatment. There was no signi cant difference in tumor size of MDA-MB-231 tumors between treated and untreated groups (Fig. 5d, g "MDA-MB-231-Treated" and "MDA-MB-231-Untreated"). The weights of the B16F10-and MDA-MB-231-xenografted mice in both groups were not signi cantly different (Fig. 5b, e). The median survivals of B16F10-xenografted suppressed, recurred, and untreated groups were 34, 14.5, and 9.5 days, respectively (p < 0.001). The median survivals of the MDA-MB-231xenografted treated and untreated groups were 29 and 30 days, respectively (p > 0.05). No mice had lung or liver metastases at the monitoring end point.

Discussion
Our research used an 808 nm laser to treat B16F10 tumors engrafted into nude mice. We explored the feasibility of 18 F-5-FPN PET imaging for PTT response evaluation, further compared with the detection of in ammation and other tumors to ascertain its speci city for MM. 18 F-5-FPN is a promising probe to assess the therapeutic response of MM and is likely to be a new strategy for the clinical response evaluation and follow-up of MM. 18 F-5-FPN and 18 F-FDG PET acquisitions were performed on B16F10-engrafted models in treated and untreated groups at each time point (Fig. 2a). In the treated group (Fig. 2c), tracer accumulation in tumors after PTT (Day 1, 2) declined from baseline. No recurrence was discovered via gross inspection within 6 days after treatment. Nevertheless, regional uptakes were observed clearly on Day 5 and 6 in mice. Then the tumor relapsed with corresponding tracer accumulation. Hence, 18 F-5-FPN PET imaging might help to reveal an occult residual recurrence before observation.

Tumor cells were killed by PTT and tumor size shrank. Residual tumor cells regrew and tumor relapsed
later. In the quantitative analysis of images (Fig. 2g, h and Table 1), the mean tumor uptakes of 18 F-5-FPN in the treated group declined on Day 2 and 6, then enhanced, which represented the diminution of melanin in tumor tissue after PTT followed by escalation. Serial 18 F-5-FPN PET imaging recorded the course accurately with high sensitivity. No difference was found for 18 F-FDG uptake before and on Day 1 after therapy, implying less sensitivity compared to 18 F-5-FPN. Therefore, 18 F-5-FPN might have a higher sensitivity to an early response.
In the untreated group, tumor size was small initially with low 18 F-FDG uptake and enlarged gradually with increasing 18 F-FDG uptake, while 18 F-5-FPN values remained on a high level. Although the mean tumor uptakes attenuated due to local necrosis later, the maximum uptakes of the whole tumor were still high (Day 9, 18 F-5-FPN, 35.92 ± 5.61%ID/g. Day 8, 18 F-FDG, 29.16 ± 5.53% ID/g).
CT and MRI are generally used to evaluate therapy response in the clinic. However, they sometimes fail to discriminate between real progression and pseudoprogression. One investigation revealed that 22 out of 227 patients responded atypically, and lesions were con rmed to be in ammatory in ltration and necrosis, not tumor proliferation by biopsy [7]. Moreover, conventional techniques cannot monitor tumor response to immunotherapy in a timely and accurate fashion [6,8]. The previous article reported that the pseudoprogression rate via conventional techniques was as high as 10-15% for malignant melanoma patients treated with immunotherapy [21]. Changes in tumor function and metabolism always precede changes in size. Thus, functional imaging may facilitate the detection of early therapy response [7,22,23].
By now, a few probes have been explored for early diagnosis or staging of MM, but few potential agents have been evaluated for response evaluation, except for 18 F-FDG and 18 F-FLT [6-8, 13, 14]. 18 F-FDG PET/CT imaging plays an important role in response evaluation to conventional chemotherapy, and also has some value in targeted therapy and immunotherapy [7,8]. Geven et al. used 18 F-FDG and 18 F-FLT PET imaging to assess the response of MM models to BRAF-mutant inhibitor and revealed that tumor uptakes of 18 F-FLT had no difference before and after therapy, though the accumulation of 18 F-FDG decreased obviously after therapy [13]. Because both MM and immune-related in ammation accumulate 18 F-FDG, 18 F-FDG PET imaging may make it quite challenging to differentiate between real progression and pseudoprogression [7]. One study investigated 12 metastatic MM patients under anti-CTLA-4 antibody treatment. It concluded that the SUV max changes of 18 F-FDG and 18 F-FLT PET images would not offer reliable value to predict late response 3 months after the rst therapy [14]. However, another article concluded that early 18 F-FDG PET/CT scanning could predict the response of MM to immunotherapy [6].
According to the literature, it is still controversial whether 18 F-FDG PET imaging might estimate therapy response of MM well or not, and whether the imaging can exclude the interference of in ammation and detect an occult residual recurrence e ciently in the clinic.
Similarly, in our imaging surveillance, the area of 18 F-FDG accumulation was larger than that of 18  In the comparative imaging, only B16F10 tumor was distinctly visible on 18 F-5-FPN PET imaging, while three models were all obvious on 18 F-FDG PET images (Fig. 4a-c). Quantitative analysis also corroborated that B16F10 tumor accumulated considerable 18 F-5-FPN, and all three showed high uptakes of 18 F-FDG ( Fig. 4d and Table 2). Moreover, 18 F-5-FPN uptakes of in ammatory and healthy sides resembled, while in ammation accumulated much more 18 F-FDG than normal muscle. 18 F-5-FPN displays an excellent a nity for melanin, but the speci city of 18 F-FDG is not good enough to distinguish between tumor and in ammation, nor between pigmented tumors and other carcinomas. Local in ammation may emerge after PTT, and 18 F-FDG cannot reveal early response to therapy accurately. This may be correlated to the nding that 18 F-5-FPN uptake declined early after PTT, while a reduction in 18 F-FDG uptake was not obvious in this research (Fig. 2g, h). Here, the smaller size of the B16F10 tumor may be responsible for its lower uptake of 18 F-FDG than that of the other two models. This can also be demonstrated by the value > 13%ID/g over the increasing volume of B16F10 tumor in the imaging surveillance.
The therapy landscape of MM has made profound progress and several drugs for targeted therapy and immunotherapy, including vemurafenib, dabrafenib, ipilimumab, pembrolizumab, and nivolumab, have been approved following clinical trials [2]. Overall, targeted therapy achieves short-term palliation, immunotherapy needs a long time to judge the response, and the two approaches induce different kinds of side effects [2][3][4][5]24]. Thereby, this research did not adopt drug therapy but PTT. Exposed to an 808 nm laser, the temperature of B16F10 tumors increased rapidly [15,[25][26][27][28], but only slightly in MDA-MB-231 tumors. Melanin can transform optical energy into heat e ciently. Histological analysis, imaging surveillance, and survival analysis identi ed that the PTT protocol inhibited tumor growth strikingly over a short time, although all tumors eventually recurred. It said that tumor cells might tolerate hyperthermia via the mediation of heat shock proteins to enhance their survival rate, which might explain why tumors relapsed after PTT [29]. Thus, several reports have explored gold nanomaterials or other materials to improve the e cacy of PTT and even combined PTT with chemotherapy or other treatment strategies [15,[25][26][27][30][31][32][33][34][35].