The mechanism of the excessive production of hormones by many endocrine neoplasms is not completely understood in most instances. Obviously, alterations in the levels of transcription factors, receptors, or perturbations in signal transduction pathways might explain a number of the examples. It is unlikely that the evolving karyotypic instability of the stage of progression is directly involved in most examples, since the most common examples of functioning endocrine neoplasms are in adenomas and in highly differentiated carcinomas that are likely to be very early in the stage of progression. However, several investigations have demonstrated specific mutational alterations in receptors and components of signal transduction pathways that are probably responsible for the excessive specific hormone production in the individual cases described.
In 1990 Lyons and associates (Lyons et al., 1990) described a putative cellular oncogene termed gsp. The wild-type protooncogene of gsp is the α-subunit of the G protein involved in the response of pituitary cells to growth hormone–releasing hormone (GHRH). A diagram of the effect of the mutations that have been found in this proto-oncogene is seen in Figure 18.7. The amino acid substitutions noted result in constitutive activation of adenyl cyclase (AC) and subse- quent cyclic AMP formation, which, in turn, activates a signal transduction pathway leading to both hyperplasia and hyperfunction (cf. Spada et al., 1994, 1998).
Figure 18.5 Structure and organization of different pre-pro hormones expressed in the pancreas. The numbers at the upper right of each structure indicate the number of amino acid residues. Pre-pro insulin is an example of a precursor from which the entire sequence is used for synthesis of one bioactive peptide. Pre-pro glucagon and pre-pro VIP (vasoactive intestinal peptide) are polyprotein precursors containing sev- eral homologous but not identical bioactive peptides. Pre-pro TRH (thyrotropin releasing hormone) is an example of a polyprotein precursor containing five identical copies of the bioactive peptide. (Adapted from Rehfeld et al., 1996, with permission of the authors and publisher.)
Figure 18.6 Distribution of some peptide hormone-producing tissues arising from the neural crest and belonging to the APUD system. The associated neoplasms of these tissues are designated by boxes. (After Schein, 1973, with permission of the author and publisher.)
At least one of these mutations has been found in about 40% of growth hormone–secreting pituitary neoplasms in several series (Johnson et al., 1999; cf. Spada et al., 1998). However, an- other series from Australia (Boothroyd et al., 1995) found a much lower incidence of such muta- tions in pituitary and other neoplasms. A second cellular oncogene, gip2, has mutations in similar regions of an α subunit of a different G protein whose function results in decreased cy- clic AMP levels and decreased calcium ion mobilization (cf. Spada et al., 1992). The gip2 cellu- lar oncogene has been identified in about 10% of nonfunctioning pituitary adenomas as well as in neoplasms of the ovary and adrenal cortex. However, the exact functional alteration contribut- ing to the neoplastic process or alterations in function in these neoplasms is not absolutely clear (cf. Spada et al., 1998).
Another example of mutations in a proto-oncogene leading to the formation of a cellular oncogene is seen with the thyrotropin receptor in adenomas of the thyroid gland. In this instance, the multimembrane domain receptor has been shown to be mutated in from 50% (Krohn et al.,1998) to 80% (Parma et al., 1995) of hyperfunctioning thyroid adenomas. Figure 18.8 gives a diagram of the thyrotropin receptor, indicating in black the amino acids where mutations were demonstrated (Parma et al., 1995). Mutation in the receptor appears to cause hyperfunction and proliferation of the adenomatous thyroid epithelium by constitutively elevating cyclic AMP (Krohn et al., 1998; Parma et al., 1995). About 30% of hyperfunctioning thyroid adenomas also exhibit the presence of the gsp cellular oncogene (O’Sullivan et al., 1991; Suarez et al., 1991).
In contrast to mutations in the G proteins in endocrine neoplasms noted above, mutations in N-ras, H-ras, and K-ras are relatively infrequent in thyroid and adrenal neoplasms in the hu- man (Moley et al., 1991; Moul et al., 1993). Only about a quarter of adenomas and carcinomas of the thyroid exhibit ras gene mutations (Namba et al., 1990). Interestingly, Shi et al. (1991) reported that a 50% to 80% rate of ras codon 61 mutations occurred in thyroid tumors develop- ing in an iodide-deficient area. While adrenocortical neoplasms exhibited low levels of G-pro-
Figure 18.7 Diagram of the effect of single amino acid substitutions replacing Arg 201 with either Cys or His, or Gln 227 with either Arg or Leu in the protooncogene α-subunit of Gs, a G protein. The substitu- tions lead to the constitutive activation of adenyl cyclase (AC) and cyclic AMP formation by inhibiting GTPase activity. Increased cyclic AMP levels activate protein kinase A, which in turn phosphorylates CREB, initiating gene transcription. (Adapted from Spada et al., 1994, with permission of the authors and publisher.)
tein mutations, they were found to overexpress insulin growth factor-II in more than 80% of carcinomas studied (Gicquel et al., 1995).
One may speculate that the mutations seen in the G proteins and receptors in pituitary and thyroid adenomas reflect initiating mutations, since they appear to function in a dominant man- ner and would likely produce effects in the cell even if only one allele were mutated. However, considerable further studies will be required before such questions can be answered.
Effects on the Host-Tumor Relationship Produced by Hormones Elaborated by Neoplasms of “Nonendocrine Tissues”
Ectopic hormone production, a term coined by Liddle and associates (1962), refers to the pro- duction of hormones by neoplasms of nonendocrine origin or the production of hormones by endocrine neoplasms not associated with the normal gland of origin (cf. Smith, 1975). It is also possible that the “ectopic” production of a hormone by a neoplasm is the result of a relatively small number of cells in the tumor that are derived from normal endocrine tissue and produce the hormone in question. Under these circumstances, such hormone production by the neoplastic tissue cannot be considered as truly ectopic. In fact, a number of criteria have been delineated, one or more of which must be satisfied in order to demonstrate that the neoplasm is producing the hormone inappropriately or ectopically (cf. Wajchenberg et al., 1994; Baylin and Mendel- sohn, 1980). These characteristics, which confirm that a given neoplasm is the source of ectopic hormone production, are listed below. There have been suggestions that all neoplasms produce
Figure 18.8 Diagram of the thyrotropin receptor structure in the membrane of thyroid epithelial cells. The circles with letters indicate the amino acids in which mutations have been found. The black symbols refer to mutations found in a series of adenomas for which the entire exon 10 of the protein was sequenced (Parma et al., 1995), while the white circles refer to other mutations found in toxic adenomas of the thyroid. (Reproduced from Parma et al., 1995, with permission of the authors and publisher.)
at least one hormonally active substance ectopically (Odell et al., 1977). On this basis as well as other studies, it is likely that many of the unexplained changes seen in the tumor-bearing host may be the result of one or more known or unknown hormones being elaborated by neoplasms.
1. Abnormal endocrine function
2. Disappearance of endocrine abnormalities following removal of neoplasms
3. Persistence of elevated levels of hormone following removal of gland normally pro- ducing the hormone
4. Increased concentrations of the hormone in the neoplasm compared with surrounding tissues
5. Molecular evidence for the synthesis of the hormone by the neoplastic cells
The first recognition that hormones elaborated by neoplasms arose from cells not consid- ered to elaborate such hormones was described by Brown in 1928. More than 30 years later, Liddle et al. (1969) recognized that lung cancers were associated with Cushing syndrome, nor- mally seen with hyperadrenocorticism, and demonstrated that the neoplasms contained large amounts of biologically active adrenocorticotropic hormone (ACTH). While originally consid- ered relatively rare, Cushing syndrome resulting from ectopic production of ACTH-like polypeptides by lung and other neoplasms is increasingly being recognized such that a wide va- riety of neoplasms may be associated with “ectopic” Cushing syndrome (cf. Odell, 1997). Some of these neoplasms produce a hormone that is biologically, physically, chemically, and immuno- logically indistinguishable from human pituitary adrenocorticotropic hormone. Others produce
variously modified forms of the hormone, as discussed below. Through its action on the adrenal glands, this ectopically produced hormone may cause all of the symptomatology of Cushing syndrome, a clinical picture of moon face, enlarged abdomen with linear streaking of the skin of the lower trunk, hypertension, and salt retention. The production of such a hormone is not sup- pressible by the administration of corticosteroid hormones, natural or synthetic.
Table 18.4 lists a variety of neoplasms derived from nonendocrine tissues that elaborate one or more hormone-like substances resulting in specific clinical findings. As noted in the ta- ble, several neoplasms—e.g., small-cell carcinoma of lung, renal cell carcinoma, adrenal neo- plasms—may give rise to more than one hormone produced ectopically. In fact, lung cancers may elaborate more than 20 different hormones, most of which are not elaborated by pulmonary tissue in the normal state (Keffer, 1996). While ectopic hormone production is relatively com- mon in some neoplasms, such as small-cell carcinoma of the lung, and contributes to the major- ity of the clinical symptomatology seen in general regardless of the source of the hormone, some ectopic hormone production is relatively rare, such as prolactin and rennin (cf. Odell, 1997).
Table 18.4 Some Ectopic Hormone–Producing Neoplasms and Their Effectsa