Richard P. Baum
Department of Nuclear Medicine, Center for PET/CT, Zentralklinik Bad Berka, Germany
The first aim of THERANOSTICS is the molecular imaging and accurate diagnosis of a disease. This can be achieved by using a specific vector (e.g. G protein coupled peptides like the somatostatin analogues or antibodies) - the KEY - targeting specific ligands - the LOCK - e.g. somatostatin receptors or antigens); an easily available generator-derived diagnostic trivalent radiometal with convenient labeling characteristics like Ga-68 can then be attached by a linker to the KEY which enables the visualization of specific molecular properties of a cell (e.g. of cell surface or intracellular structures). The single most imperative aspect of PET/CT is its ability to quantify a disease at the molecular level. The second goal - personalized treatment - is achieved by utilizing the same molecular vectors as used for imaging and by substituting the generator-derived radionuclide by a therapeutic radionuclide (e.g. applying beta emitters like Lu-117 or Y-90). Thirdly, individualized patient dosimetry can be performed by using a longer lived PET radioisotope before therapy and serial PET scanning or (with less accuracy, but more practically) by post-therapeutic dosimetry using the administered therapeutic radioisotope and taking advantage of the gamma emission of radiotherapeutics like Lu-177. Finally, post-therapeutic follow-up and therapy control can be performed using again quantitative PET/CT and generator-derived positron emitters.
Today, the most notable example are neuroendocrine tumors (NET) - imaged and treated with radiolabeled peptides forming a diagnostic-therapeutic pair like Ga-68/Lu-177 labeled somatostatin (SMS) analogues. Another emerging application is the treatment of bone metastases - imaged with bisphosphonate-based agents (e.g. Ga-68 BPAMD) and treated by the same vector labeled with Lu-177.
The Bad Berka experience: At the Neuroendocrine Tumor Center Bad Berka (which has been certified as ENETS Center of Excellence in March 2011) we are working extensively in the field of diagnosis and treatment of NET in a multidisciplinary team, especially over the last 10 years (in Germany, the first patient was treated by us in July 1997). Over 3,000 treatment cycles of Peptide Receptor Radionuclide Therapy in nearly 1000 patients have been performed up to now. First Y-90 and later Lu-177 (in use in our center since 2004) labeled somatostatin analogues have been administered alone or in combination to patients with progressive neuroendocrine tumors, non-responsive to other treatments like surgery, octreotide/interferon treatment, kinase inhibitors or chemotherapy (we are using mainly DOTATATE and DOTATOC and have administered also Lu-177 DOTANOC to a few patients). Lu-177 DOTATATE or -DOTATOC are predominantly used for smaller metastases or in patients with impaired renal or hematological function. Our group was the first to systematically administer Lu-177 and Y-90 consecutively in the same patient (DUO-PRRNT) and to concurrently apply PRRNT with Y-90 and Lu-177 peptides on the same day (TANDEM PRRNT). We were also among the first to administer (in 1999) intra-arterial PRRNT using Y-90 (IA-PRRNT) for enhancing the delivery to large, inoperable primary tumors and bone and hepatic metastases. We also inaugurated the combined use of internal (PRRNT) and external (EBRT) radiotherapy (COMBIERT) for the treatment of neuroendocrine tumors (e.g., bone invading glomus tumors, progressive meningiomas as well as large mediastinal tumors or pleural carcinosis or extensive bone metastases).
In all patients, Ga-68 DOTA-SMS analogue receptor PET/CT is performed before and after PRRNT (over 5,700 Ga-68 DOTA-SMS PET/CT studies have been performed until now). The decision for administering PRRNT is undertaken based on high Somatostatin receptor (SSTR) expression. We have extensive experience with imaging (and therapy) using the three somatostatin analogues DOTATATE, DOTATOC and DOTANOC, and recently had the chance of using a Ga-68 labeled SSTR antagonist first time in humans (JR10, kindly provided by Helmut Mäcke).
Analysis of long term follow-up after PRRNT has been performed in 454 patients (mean age 59.1 years, 248 male, 206 female), who received a total of 1,303 treatment cycles (mean activity 4.1 GBq, min. 1.05 GBq, max. 7.5 GBq per cycle, time between cycles 3 to 6 months) The number of patients/cycles are: 111 pts/1 cycle; 106/2; 74/3; 75/4; 42/5; 23/6; 9/7; 1/8. For kidney protection, patients were well hydrated and received an L-lysine/L-arginine solution infused intravenously for 4 hours beginning 30 minutes before PRRT. In addition, patients treated with Y-90 SMS analogues are pretreated with GELOFUSIN since 2008 (overall we have experience in using this nephroprotecting agent in over 1,500 treatment courses). Before each new treatment cycle, restaging was performed by morphologic (CT/MRI) and molecular imaging (Ga-68 DOTA-SMS PET/CT, in selected cases also FDG or fluoride PET/CT were performed), and blood chemistry and tumor markers (CgA, serotonine, specific hormones) were determined. Renal function was serially determined by Tc-99m MAG3 scan/clearance (TER) and by Tc-99m DTPA (GFR) measurements. Tumor and normal organ dosimetry (MIRD/OLINDA) was performed after PRRNT (using Lu-177-DOTA-SMS analogues). All data of patients treated by PRRNT are entered in a structured ACCESS database (284 items /patient) for prospective evaluation since the very beginning.
Tumor response in patients with NETs of non-pancreatic origin and pancreatic NET (pNET) after a mean follow-up of 2 years was as follows: Complete remission (CR), partial remission (PR), minor response (MR) after 3 cycles (progressive disease before) was seen in 52% of patients with pNET (48% in other non pancreatic NET); diseased was stabilized in 39% of pNET as compared to 45% in non pancreatic NET. 36 patients with advanced disease died of PD. Objective tumor responses (including improvement of clinical symptoms) were seen in 93% (91% pNET) of the patients. Significant hematological toxicity (mainly erythrocytopenia, rarely neutropenia, and severe thrombocytopenia) occurred in less than 15% of all patients. MDS developed in 5 patients (all of them received also chemotherapy before). End stage renal insufficiency was not observed in any of the patients with normal kidney function before PRRT. In most patients receiving Lu-TATE alone (n=417 cycles), serum creatinine and TER/GFR did not change. Therefore, the probability and magnitude of renal toxicity can be significantly reduced when PRRT is administered in fractionated doses in patients without any preexisting risk factors and under appropriate nephroprotection. As risk factors we identified chemotherapy, diabetes mellitus, hypertension, Hedinger's syndrome, and cachexia.
We also have been treating patients with a single functional kidney (24 patients) and 3 patients on hemodialysis due to renal insufficiency (worldwide first experience in 2009) with fractionated low dose PRRNT. None of these patients (with still functioning single kidney) showed grade 3 or 4 nephrotoxicity. PRRNT resulted in PR in 36% and stable disease (SD) in 36% of the pts, 28% had PD. 14 had grade 1 erythrocytopenia, 3 grade 1 leukocytopenia and 3 had grade 1 thrombocytopenia. No significant hematotoxicity was observed in the three patients on dialysis.
Thus, in patients with progressive neuroendocrine tumors, fractionated, personalized PRRNT with lower doses of radioactivity given over a longer period of time (Bad Berka Concept) results in good therapeutic responses and severe hematologic and/or renal toxicity can be avoided. Quality of life of the patients can be highly improved. A recent analysis of 416 patients (all NET subtypes) treated at the Bad Berka Neuroendocrine Tumor Center showed a median overall survival from the time of first diagnosis of 210 months and a median survival after 1st PRRT of 59 months (published Rotterdam data 46 months). Our experience confirms a previous report that - compared with historical controls - there is a benefit in overall survival from the time of diagnosis of several years.
Following the concept of THERANOSTICS, we have been using clinically a number of other peptides and radiopharmaceuticals for both diagnosis and therapy of tumors (most of these as "first in humans"). Very promising is the bisphosphonate-based agent BPAMD. PET/CT with Ga-68 DOTA-BPAMD is more sensitive and results in much a high resolution images as compared to the conventional bone scan using Tc-99m MDP. We have treated meanwhile >10 patients with widespread, painful skeletal metastases, presenting with progressive disease and refractory to conventional treatment administering Lu-177 BPAMD. Dosimetry shows that due to the long half life of the radiopharmaceutical in the metastases (>80 hours), the tumor doses delivered were quite high (ranging from 2.4 - 209 mGy/MBq, the wide range is due to the different size of the lesions). A significant reduction in osteoblastic activity of the bone metastases was seen on the follow up PET/CT. The treatment was very well tolerated by all patients without any significant adverse effects. There were only minor changes in blood cell counts (not needing any intervention), and no significant alterations of serum creatinine/BUN or other lab parameters were observed.
In 2009, our group was pioneering the use of a positron emitter labeled gastrin-releasing Peptide (GRP) selective bombesin antagonist for GRP receptor PET/CT imaging of metastatic breast, lung and prostate cancers using Ga- 68 Demobesin. The first ever Lu-177 Demobesin therapy was also administered in 2009 in a patient with metastatic prostate cancer.
Many other Ga-68 labeled compounds have been first used - gastrin for NET, the GRP-R agonist AMBA for imaging (and therapy) of cancers of the breast, lung and prostate, Ga-68 DOTA Tyrosine for imaging brain tumors, Ga-68 MAA for lung perfusion scintigraphy, Ga-68 alpha-MSH (melanocyte stimulating hormone) in a patient with choroid melanoma and Ga-68 labeled glucose (without success) aimed for determining the metabolic status of tumors. This essentially underlines the emergence of Ga-68 as an effective radionuclide for high resolution PET imaging (the author named it in 2004 “the Tc-99m for PET”).
Novel therapeutic approaches include the first antiangiogenesis therapy using Lu-177 RGD peptide and I-131 phenylalanine endoradiotherapy of brain tumors. The first human study using the longer-lived (3.93 h half-life), generator-derived (Ti-44/Sc-44) trivalent metallic positron emitter Sc-44 (Scandium-44 obtained from a Ti-44/Sc-44 generator with a very long half life of Titanium-44 of 59.2 years) DOTATOC was performed in 2009. This should definitely inspire the development of new longer-liver PET radioisotopes in order to cover longer imaging periods (which is of special interest for predictive, personalized dosimetry). The objectives of THERANOSTICS could be achieved by the additional use of Sc-47, a beta emitter, for therapy.
Looking into the future - towards personalized medicine: It is worth emphasizing again that every patient is different not only when it comes to be able to tolerate a particular therapy or the degree to which a patient responds, but also concerning diagnostics. Hence, it becomes imperative to choose the appropriate modality for THERANOSTICS. In this context, as we move towards personalized medicine, we must first improve the diagnostic information obtained from PET/CT, i.e., by adequate quantification.
We have developed the Bad Berka Molecular Imaging Tool (BBQ-MIT), an automatic user-independent routine for segregation and quantification (e.g. molecular tumor volume - MTV) of receptor positive neoplastic lesions detected by molecular PET/CT using Cognition Network Language (CNL) and software package provided by Definiens. This prototype routine built on CNL for DICOM PET/CT images enables the automatic analysis and quantification of lesions. It seems especially promising for shortening the time needed to evaluate a PET/CT scan showing many lesions, and will improve reproducibility as well as increase sensitivity of lesion detection. BBQ-MIT seems to be a quantum step forward towards fast and accurate analysis of serial PET/CT studies, allowing to assess tumor response early in the course of therapy and monitoring further follow-up, thus enabling effective personalized patient management.
Another very important future aspect of personalized medicine and THERANOSTICS is dosimetry, which is the single most important factor to ensure maximum dose delivered to the tumor and has decisive influence on the treatment dose/activity. Newer protocols for 3D dosimetry are developed and must be applied clinically to ensure a more individualized and accurate treatment approach. Dose assessment and predicting the molecular response on the basis of pre-therapeutic PET/CT studies is the final goal which can be accomplished by the use of longer lived PET tracers for dosimetry.