Diagnosis and Management of Pituitary Tumors

302 QUABBE AND PLÖCKINGER(VIP) (171) or peptide histidine-methionine (172), and even toglucose (173), as well as paradoxical suppression by stimulatoryagents, dexamethason (174) and galanin (175), have also beenreported. However, the paradoxical GH response is not specificfor acromegaly. It may also occur in diabetes mellitus, hepatic orrenal failure, hypothyroidism, and other diseases unrelated toacromegaly, e.g., schizophrenia (166,167,176). Paradoxical GHresponses are not a diagnostic tool for acromegaly.On the other hand, it has been proposed that the GH responseto TRH predicts the GH-suppressive effect of bromocriptine treatmentand that the GH increase following VIP or GnRH administrationpredicts a smaller GH response to bromocriptine (171).However, the predictive value for individual patients is low, andthe response to dopamine-agonist treatment can be better tested byacute administration of the drug.The GH response to TRH has also been evaluated as a possibleprognostic parameter following surgery. A preoperative paradoxicalGH response as well as its disappearance following surgeryhave been suggested to predict a good prognosis (168,177). However,the correlation has not been confirmed by others (169,178).It has also been postulated that the TRH response is more oftenpositive in patients with PRL-cosecreting somatotropinomas thanin those with pure GH-secreting adenomas (179,180), but this hasnot been confirmed by others (181). Some of the differences maybe owing to the lack of strict criteria for a specific response interms of both its magnitude and its timing. These are important inview of the occurrence of spontaneous secretory GH episodesin many patients (182).The mechanism responsible for the paradoxical response is atpresent unknown. Removal of the hypothalamic influence on thepituitary gland (e.g., by autotransplantation of the pituitary glandunder the kidney capsule in experimental animals) induces a permanentparadoxical response to TRH (183). Alternatively, theparadoxical response may be owing to the (pathological) expressionof specific receptors for TRH and/or GnRH on the somatotrophadenoma (184).GHRH TEST The GH response to the hypothalamic GHRHis preserved to some degree in patients with acromegaly. Attemptshave been made to use this response for prognostic purposes. Thus,an exaggerated GH response to GHRH correlated with a nonsuppressibleor paradoxically increasing GH during an oral glucosetolerance test (oGTT). It was suggested that this was indicative ofa more active disease (185). In other studies, the predictive valueof the GH response to GHRH for the success of bromocriptinetreatment was evaluated. The response to bromocriptine was betterin patients with a smaller GH reaction to GHRH (180,186).However, partly in view of large interindividual variations, theseobservations have not found a place in the treatment decisions forpatients with acromegaly.The synthetic GH-releasing peptide (GHRP, a hexapeptide)and related compounds stimulate GH secretion in most patientswith acromegaly (187–189). However, although the syntheticGHRPs hold promise for the treatment of patients with GH deficiencyof hypothalamic origin, at present they have no diagnosticor prognostic value for patients with acromegaly.SCINTIGRAPHIC TUMOR VISUALIZATIONScintigraphic methods for in vivo visualization of somatostatinreceptors are now available. Somatostatin receptors have beendemonstrated in surgical specimens of GH-secreting adenomas byautoradiography (190) and also in vivo (191–193). Early experiencewas gained with 123 I-Tyr 3 -octreotide (191,192). Subsequently,the 111 Indium-labeled derivative of the octreotide analog,111 In-DTPA-D-Phe-octreotide (Octreoscan ® ), became commerciallyavailable. A good correlation between scintigraphic demonstrationof somatostatin receptors on the adenoma and subsequentresponse to octreotide treatment was reported with the 123 I-Tyr 3 -octreotide preparations (191,192). Our own experience with theOctreoscan procedure is less encouraging. Since 111 In-DTPA-D-Phe-octreotide visualizes the normal pituitary gland more frequentlythan 123 I-Tyr 3 -octreotide (193), it becomes difficult todistinguish an adenoma with a low number of receptors from anormal gland (194). Moreover, in our experience, the scintigraphicresults were not correlated with the GH response to octreotidetreatment, nor did they predict a volume reduction in response topreoperative octreotide therapy (194). Thus, although theOctreoscan is apparently of great help in the localization ofgastroenteropancreatic tumors (195), its value for the diagnosis ofa GH-secreting pituitary adenoma and for treatment decision is atpresent questionable. The Octreoscan procedure delivers a relativelyhigh radiation dose of approx 16 mSV (193).Other radioactively labeled octreotide derivatives are beingdeveloped with the use of gallium isotopes for gamma-scintigraphyand also for positron emission tomography (PET; 196,197). Technetium-99m-pentavalent-dimercaptosuccinicacid (99mTc[V]DSMA)imaging has also been used to visualize pituitary tumors, independentof their hormonal activity or their somatostatin receptor status(198). A PET study has been performed with fluorine-18-2-fluorodeoxyglucose (FDG; 199). This procedure measures themetabolic activity of a pituitary adenoma and not its hormonalactivity. PET has also been used for demonstration of dopaminereceptors on pituitary tumors, including somatotroph adenomas.Two different dopamine D 2 receptor ligands were used in anattempt to correlate receptor demonstration with the result ofdopamine agonist treatment (200). Dopamine D 2 receptors havealso been visualized scintigraphically in prolactinoma patientsusing iodobenzamide in order to predict treatment response todopamine agonist drugs (201). Patients with somatotroph tumorsand their response to dopamine agonists have yet to be studied.The clinical importance of these developments for the diagnosisand prognostic predictions in patients with acromegaly awaitsfurther investigation. PET, in particular, is not widely availableand remains largely a research tool at the present time.HYPERPROLACTINEMIA,AND MULTIHORMONAL ADENOMASElevated PRL concentrations are somewhat common inacromegaly and occur in about one third of the patients (202,203).Hyperprolactinemia may be owing to cosecretion of PRL from thetumor or to pituitary stalk compression by a macroadenoma. Stalkcompression interrupts the dopamine flow from the hypothalamusthat inhibits PRL secretion from the nonadenomatous pituitarygland. PRL concentrations in such patients usually remain belowapprox 200 µg/L. In other patients, hyperprolactinemia is owing toPRL secretion from the pituitary adenoma. PRL is then eithersecreted from the same cell that produces GH, or each hormone issecreted from its own cell (mammosomatotroph cell and mixedGH/PRL adenoma, respectively [115]). Cosecretion of TSH withGH resulting in acromegaly with concomitant secondary hyperthyroidismdoes occur, but is a very rare combination (115,204,205).