Staging of Pancreatic Carcinoma by FDG-PET
3.5.1 T Stage
The determination of the T stage is limited by the spatial resolution of PET having poor anatomical orientation. Although the introduction of PET/CT improves the topographical focus assign¬ment in pancreatic cancer, it is doubtful that the determination of, e.g., infiltration will surpass that of multislice CT or MRI alone, as the true spatial extent of a tumor is difficult to assess by metabolic imaging and the "size" of the focus is largely dependent on the window/threshold cho-sen (Ruf et al. 2006).
3.5.2 N Stage
The detection and assessment of lymph node metastases is limited in both FDG-PET and con-ventional imaging.
In our own analysis of 53 patients, FDG-PET achieved a sensitivity of 63% and a specificity of 93%, a result comparable to the data in the litera-ture, which reports a sensitivity of 41%-71% and a specificity of 63%-100% (Higashi et al. 2003; Pakzad et al. 2006). The low sensitivity is partially explained by partial volume effects in small lym¬phatic lesions that cause a lesser signal than the primary. Moreover, the focus of a peripancreatic lymph node may appear "merged" with that of the pancreatic primary, making the classification as separated lesions difficult unless morphologi¬cal information is at hand.
3.5.3 M Stage
As the exclusion of distant metastases is crucial for the performance of curative surgery, FDG-PET has a decisive advantage due to the whole-body technique (Fig. 3.2).
Most frequently metastases to the lung, the bone marrow, and especially the liver are encoun-tered. Frohlich and coworkers (1999) reported in a preoperative study on 168 patients a sensitivity of 68% for the detection liver metastases. Their subanalysis revealed the difficulty of FDG-PET for the detection of small lesions (<1 cm; sensi-tivity 43%), whereas larger lesions were detected in 97% of all cases. Our own results confirm this observation with a sensitivity of 43% and 97% for metastases smaller and larger than 1 cm re-spectively. The reasons for the low sensitivity in small lesions are probably partial volume effects again, which especially in hepatic metastases may hinder diagnosis due to the physiological FDG uptake of hepatic tissue. Moreover, false positives based on cholangitic reactions in case of cholestasis may also occur.
Despite these limitations, an impact of up to 40% on the decision of the therapeutic strategy has been reported (Higashi et al. 2003; Delbeke
et al. 1999).
3.6 Assessment of Prognosis
In addition to established prognostic factors for pancreatic carcinoma such as tumor stage or the height of the CA 19-9 tumor marker level, the information gained on tumor biology by meta-bolic FDG-PET imaging appears to deliver fur¬ther information for the assessment of survival. Detailed information of the prognostic relevance of FDG-PET and the assessment of response to therapy are given in Chap. 12.
3.7 Tracers Beyond FDG
As mentioned earlier, FDG is the most common but not the only PET tracer available. For exam-ple, in the case of pancreatic tumors with neuro-endocrine differentiation, somatostatin receptor PET, in analogy to somatostatin receptor scin-tigraphy, has become a powerful tool for the as¬sessment of disease extent (Kowalski et al. 2003). Although the initial attempts to visualize the adenocarcinoma of the pancreas by C11-acetate (synthesis of membrane components) (Rasmus-sen et al. 2004) or F18-FET (amino acid metabo¬lism) (Pauleit et al. 2005) did not prove superior to FDG, the multitude of possible combinations of ligands and positron emitting nuclides indi¬cate the potential of PET for the assessment of tumor pathobiology, as it comes closest to the ideal of true molecular imaging. However, al¬though PET imaging does allow for an accurate depiction, e.g., gene transfection both in vitro and in vivo (Penuelas et al. 2005; Gambhir et al. 1998), its usefulness will be linked to the clinical realization of such therapy concepts.
3.8 Image Fusion and PET/CT
Tracer research is not the only rapidly growing field in PET imaging. With the introduction of improved hardware (e.g., new detector crystals) and software allowing for a new generation of tomographs with a higher sensitivity and bet¬ter spatial resolution, the recent breakthrough of nuclear medicine imaging is the integration of functional images gained by PET and mor-phological images gained by CT or MRI into one image as opposed to classical side-by-side analysis (Table 3.2). This so-called image fusion was initially based on software techniques, which retrospectively generate a single data set of the separately acquired volume data of PET and CT/ MRI, ready for fused visualization. Most works in this field concerning abdominal imaging have concentrated on algorithms based on "mutual information" software programs that perform a voxel-by-voxel comparison of both volume data sets in order to generate a congruent overlay of both examinations. Our group as well evalu¬ated the image fusion of CT and FDG-PET in 102 patients with suspected pancreatic cancer with the help of such an algorithm. In 96.2% of all patients, image fusion was technically suc¬cessful and the data generated were evaluable. The overall detection rate for pancreatic cancer in the fused PET/CT images was 89.1%, making it superior to the single interpretation of either
CT (76.6%) or FDG-PET (84.4%) (Lemke et al. 2004). Another study compared PET/MR-fused images to the standard side-by-side analysis in 32 patients, which resulted in an improved topo-graphical assignment and interpretation of PET foci in 28% of all foci. On the downside, however, it must be noted that due to multiple metastases in those patients, the actual impact of improved focus assignment on therapy was limited (Ruf
et al. 2006).
This tendency for integrated imaging took a big step at the end of the 1990s, when the first PET/CT scanners were developed. The acquisi¬tion of both data sets is performed in one imag¬ing session, which allows for a direct overlay of both data sets using the respective coordinate system based on the position of the examination table. As a consequence of the almost simultane¬ous data acquisition, the technical success rate of image fusion is very high, as motion and/or positioning artifacts encountered in fusion of separately acquired sets of data can be greatly reduced. Moreover, the attenuation correction of the PET-emission scan is now performed with the rapidly acquired CT-data instead of the con-ventional, more time-consuming transmission by external Ge68-rod sources. This implies more patient-friendly examination times (approx. 40 instead of 80 min).
The integration of PET and modern mul-tislice CT combines the advantages of both ex-aminations and already has great influence espe-cially in the field of oncology (Beyer et al. 2000; Townsend 2001).
Up to the present, only one study concerning the use of integrated PET/CT for the detection of pancreatic cancer has been published (Heinrich et al. 2005). In their examination of 59 patients with suspected pancreatic cancer, the group also addressed the impact of PET/CT on patient man-agement and its cost-efficiency. Despite the fact that the CT component was only performed in a low-dose technique without contrast enhance-ment, curative surgery was abandoned in 16% of all patients as PET/CT detected previously un-known metastases. In analogy to the calculations of a preoperative PET trial for lung cancer pa-tients (van Tinteren et al. 2002), the examination itself was even cost-effective (saving US $1.066 per patient) due to the avoidance of noncurative surgery.
On the basis of these preliminary results, the use of a true diagnostic contrast-enhanced PET/
CT in pancreatic cancer patients for the selection of patients that will profit from surgery is prom-ising.
3.9 Summary
PET has become an indispensable tool for im-aging in oncology. Although the tracer FDG by itself has some limitations for the differentiation of pancreatitis from pancreatic carcinoma, its metabolic information in combination with the anatomical data from MRI or CT either by image fusion or hybrid-devices allows an extensive and complete assessment of the tumor. Apart from mere tumor visualization, the great potential of PET lies on the assessment of metabolic charac-teristics of the tumor, which in turn could pro¬vide useful information for a tailor-made tumor therapy.
3.5.1 T Stage
The determination of the T stage is limited by the spatial resolution of PET having poor anatomical orientation. Although the introduction of PET/CT improves the topographical focus assign¬ment in pancreatic cancer, it is doubtful that the determination of, e.g., infiltration will surpass that of multislice CT or MRI alone, as the true spatial extent of a tumor is difficult to assess by metabolic imaging and the "size" of the focus is largely dependent on the window/threshold cho-sen (Ruf et al. 2006).
3.5.2 N Stage
The detection and assessment of lymph node metastases is limited in both FDG-PET and con-ventional imaging.
In our own analysis of 53 patients, FDG-PET achieved a sensitivity of 63% and a specificity of 93%, a result comparable to the data in the litera-ture, which reports a sensitivity of 41%-71% and a specificity of 63%-100% (Higashi et al. 2003; Pakzad et al. 2006). The low sensitivity is partially explained by partial volume effects in small lym¬phatic lesions that cause a lesser signal than the primary. Moreover, the focus of a peripancreatic lymph node may appear "merged" with that of the pancreatic primary, making the classification as separated lesions difficult unless morphologi¬cal information is at hand.
3.5.3 M Stage
As the exclusion of distant metastases is crucial for the performance of curative surgery, FDG-PET has a decisive advantage due to the whole-body technique (Fig. 3.2).
Most frequently metastases to the lung, the bone marrow, and especially the liver are encoun-tered. Frohlich and coworkers (1999) reported in a preoperative study on 168 patients a sensitivity of 68% for the detection liver metastases. Their subanalysis revealed the difficulty of FDG-PET for the detection of small lesions (<1 cm; sensi-tivity 43%), whereas larger lesions were detected in 97% of all cases. Our own results confirm this observation with a sensitivity of 43% and 97% for metastases smaller and larger than 1 cm re-spectively. The reasons for the low sensitivity in small lesions are probably partial volume effects again, which especially in hepatic metastases may hinder diagnosis due to the physiological FDG uptake of hepatic tissue. Moreover, false positives based on cholangitic reactions in case of cholestasis may also occur.
Despite these limitations, an impact of up to 40% on the decision of the therapeutic strategy has been reported (Higashi et al. 2003; Delbeke
et al. 1999).
3.6 Assessment of Prognosis
In addition to established prognostic factors for pancreatic carcinoma such as tumor stage or the height of the CA 19-9 tumor marker level, the information gained on tumor biology by meta-bolic FDG-PET imaging appears to deliver fur¬ther information for the assessment of survival. Detailed information of the prognostic relevance of FDG-PET and the assessment of response to therapy are given in Chap. 12.
3.7 Tracers Beyond FDG
As mentioned earlier, FDG is the most common but not the only PET tracer available. For exam-ple, in the case of pancreatic tumors with neuro-endocrine differentiation, somatostatin receptor PET, in analogy to somatostatin receptor scin-tigraphy, has become a powerful tool for the as¬sessment of disease extent (Kowalski et al. 2003). Although the initial attempts to visualize the adenocarcinoma of the pancreas by C11-acetate (synthesis of membrane components) (Rasmus-sen et al. 2004) or F18-FET (amino acid metabo¬lism) (Pauleit et al. 2005) did not prove superior to FDG, the multitude of possible combinations of ligands and positron emitting nuclides indi¬cate the potential of PET for the assessment of tumor pathobiology, as it comes closest to the ideal of true molecular imaging. However, al¬though PET imaging does allow for an accurate depiction, e.g., gene transfection both in vitro and in vivo (Penuelas et al. 2005; Gambhir et al. 1998), its usefulness will be linked to the clinical realization of such therapy concepts.
3.8 Image Fusion and PET/CT
Tracer research is not the only rapidly growing field in PET imaging. With the introduction of improved hardware (e.g., new detector crystals) and software allowing for a new generation of tomographs with a higher sensitivity and bet¬ter spatial resolution, the recent breakthrough of nuclear medicine imaging is the integration of functional images gained by PET and mor-phological images gained by CT or MRI into one image as opposed to classical side-by-side analysis (Table 3.2). This so-called image fusion was initially based on software techniques, which retrospectively generate a single data set of the separately acquired volume data of PET and CT/ MRI, ready for fused visualization. Most works in this field concerning abdominal imaging have concentrated on algorithms based on "mutual information" software programs that perform a voxel-by-voxel comparison of both volume data sets in order to generate a congruent overlay of both examinations. Our group as well evalu¬ated the image fusion of CT and FDG-PET in 102 patients with suspected pancreatic cancer with the help of such an algorithm. In 96.2% of all patients, image fusion was technically suc¬cessful and the data generated were evaluable. The overall detection rate for pancreatic cancer in the fused PET/CT images was 89.1%, making it superior to the single interpretation of either
CT (76.6%) or FDG-PET (84.4%) (Lemke et al. 2004). Another study compared PET/MR-fused images to the standard side-by-side analysis in 32 patients, which resulted in an improved topo-graphical assignment and interpretation of PET foci in 28% of all foci. On the downside, however, it must be noted that due to multiple metastases in those patients, the actual impact of improved focus assignment on therapy was limited (Ruf
et al. 2006).
This tendency for integrated imaging took a big step at the end of the 1990s, when the first PET/CT scanners were developed. The acquisi¬tion of both data sets is performed in one imag¬ing session, which allows for a direct overlay of both data sets using the respective coordinate system based on the position of the examination table. As a consequence of the almost simultane¬ous data acquisition, the technical success rate of image fusion is very high, as motion and/or positioning artifacts encountered in fusion of separately acquired sets of data can be greatly reduced. Moreover, the attenuation correction of the PET-emission scan is now performed with the rapidly acquired CT-data instead of the con-ventional, more time-consuming transmission by external Ge68-rod sources. This implies more patient-friendly examination times (approx. 40 instead of 80 min).
The integration of PET and modern mul-tislice CT combines the advantages of both ex-aminations and already has great influence espe-cially in the field of oncology (Beyer et al. 2000; Townsend 2001).
Up to the present, only one study concerning the use of integrated PET/CT for the detection of pancreatic cancer has been published (Heinrich et al. 2005). In their examination of 59 patients with suspected pancreatic cancer, the group also addressed the impact of PET/CT on patient man-agement and its cost-efficiency. Despite the fact that the CT component was only performed in a low-dose technique without contrast enhance-ment, curative surgery was abandoned in 16% of all patients as PET/CT detected previously un-known metastases. In analogy to the calculations of a preoperative PET trial for lung cancer pa-tients (van Tinteren et al. 2002), the examination itself was even cost-effective (saving US $1.066 per patient) due to the avoidance of noncurative surgery.
On the basis of these preliminary results, the use of a true diagnostic contrast-enhanced PET/
CT in pancreatic cancer patients for the selection of patients that will profit from surgery is prom-ising.
3.9 Summary
PET has become an indispensable tool for im-aging in oncology. Although the tracer FDG by itself has some limitations for the differentiation of pancreatitis from pancreatic carcinoma, its metabolic information in combination with the anatomical data from MRI or CT either by image fusion or hybrid-devices allows an extensive and complete assessment of the tumor. Apart from mere tumor visualization, the great potential of PET lies on the assessment of metabolic charac-teristics of the tumor, which in turn could pro¬vide useful information for a tailor-made tumor therapy.
Комментариев нет:
Отправить комментарий