Germ Cell Tumors: The Problem
Germ cell tumors are rare neoplasms that occur most often in young adults and children. For a variety of reasons, much of what we think we understand about these tumors is highly confusing and probably wrong. Considering that these are rare tumors, you might accept a certain degree of ignorance, but sometimes the mysteries that surround rare tumors must be solved before we can make any headway understanding the more common tumors.
Also, for some strange reason, the incidence of seminomatous germ cell tumors of the testes, in the white population, has been increasing over the past 35 years (at least).
Here are the numbers, computed from the SEER (the U.S. National Cancer Institute's Surveillance Epidemiology and End Results) public use data files. The first column is the crude number of occurrences of seminomatous germ cell tumors of testes in white, non-hispanic males. The second column is the number of occurrences expressed as a proportion of all of the SEER cases for the year examined, and the third column is the number of occurrences expressed as a proportion of the U.S. population for the year examined.
crude of SEER of U.S. Pop
1973 000036 000064 000016
1974 000026 000038 000012
1975 000044 000059 000020
1976 000069 000091 000031
1977 000197 000257 000089
1978 000169 000216 000075
1979 000192 000239 000085
1980 000225 000271 000099
1981 000200 000234 000087
1982 000240 000277 000103
1983 000257 000286 000109
1984 000252 000270 000106
1985 000256 000262 000107
1986 000293 000292 000122
1987 000302 000285 000124
1988 000299 000278 000122
1989 000343 000311 000138
1990 000338 000290 000135
1991 000303 000245 000120
1992 000352 000274 000138
1993 000340 000269 000131
1994 000385 000305 000147
1995 000303 000237 000115
1996 000371 000304 000139
1997 000379 000300 000141
1998 000408 000315 000150
1999 000363 000274 000133
2000 000413 000310 000146
2001 000409 000297 000143
2002 000398 000285 000138
2003 000371 000268 000127
2004 000400 000277 000136
2005 000382 000262 000129
2006 000374 000252 000125
2007 000378 000249 000125
Here's the graph. The blue columns are the crude numbers. The maroon columns are the numbers as a proportion of the year's seer records, and the white column are the numbers as a porportion of the U.S. population in the examined year.
When the incidence of a tumor increases almost every year, and we're clueless to explain the increase, it's probably worth thinking about the problem.
Non-Seminomatous Germ Cell Tumors
There are two categories of germ cell tumors: seminomatous and non-seminomatous. The seminomatous tumors are tumors composed predominantly of a single cell type, the gonocyte. The non-neoplastic gonocyte would normally produce sperm cell in the testis. Seminomas are permitted to contain a few neoplastic trophoblasts, but otherwise, seminomas are composed of a population of large, round, monomorphic cells.
The other type of germ cell tumors is the non-seminomatous tumors, and these tumors are composed of malignant cells resembling those of the pluripotent primitive embryonic (from the early embryo) or extra-embryonic (from the placenta) malignant cells. Consequently, the non-seminomatous germ cell tumors may be teratomatous, primative embryonic, choriocarcinomatous, or some mixture of these. Can we observe the same increased incidence of non-seminomatous germ cell tumors as we saw (yesterday) in the seminomatous germ cell tumors. NO.
Here are the numbers, computed from the SEER (the U.S. National Cancer Institute's Surveillance Epidemiology and End Results) public use data files. The first column is the crude number of occurrences of non-seminomatous germ cell tumors of testes in white, non-hispanic males. The second column is the number of occurrences expressed as a proportion of all of the seer cases for the year examined, and the third column is the number of occurrences expressed as a proportion of the U.S. population for the year examined.
crude of SEER of U.S. Pop
1973 000109 000196 000051
1974 000147 000218 000068
1975 000157 000213 000072
1976 000165 000218 000075
1977 000189 000246 000085
1978 000167 000214 000075
1979 000182 000226 000080
1980 000216 000260 000095
1981 000222 000259 000096
1982 000203 000234 000087
1983 000226 000251 000096
1984 000219 000234 000092
1985 000238 000243 000100
1986 000233 000232 000097
1987 000253 000238 000104
1988 000222 000206 000090
1989 000263 000238 000106
1990 000243 000209 000097
1991 000237 000192 000094
1992 000237 000185 000092
1993 000245 000194 000095
1994 000222 000176 000085
1995 000216 000169 000082
1996 000247 000202 000093
1997 000225 000178 000084
1998 000234 000180 000086
1999 000245 000185 000089
2000 000237 000177 000084
2001 000223 000162 000078
2002 000258 000185 000089
2003 000230 000166 000079
2004 000278 000192 000094
2005 000275 000188 000092
2006 000251 000169 000084
2007 000277 000182 000091
Here's the graph. The blue columns are the crude occurrences. The maroon columns are the numbers as a proportion of the year's seer records, and the white column are the numbers as a porportion of the U.S. population in the examined year.
There's a small increase since 1973, but much of the increase is accounted for by the increase in the SEER population and the increase in the U.S. population for the same years. The relative (population adjusted) rate of occurrence of non-seminomatous germ cell tumors has not increased by much; certainly nothing like the increase seen yesterday, for the seminomatous germ cell tumors. What about the germ cell tumors that occur outside the gonads? Are they increasing in occurrence since 1973? Though germ cell tumors can occur outside the gonads, they are very rare. Here are the SEER numbers for non-seminomatous non-testicular tumors in white non-Hispanic men.
Here are the numbers for non-seminomatous non-testicular tumors in white non-hispanic males.
crude of SEER of U.S. Pop
1973 000011 000019 000005
1974 000008 000011 000003
1975 000014 000019 000006
1976 000011 000014 000005
1977 000011 000014 000004
1978 000011 000014 000004
1979 000019 000023 000008
1980 000021 000025 000009
1981 000016 000018 000006
1982 000019 000021 000008
1983 000018 000020 000007
1984 000012 000012 000005
1985 000019 000019 000007
1986 000014 000013 000005
1987 000025 000023 000010
1988 000016 000014 000006
1989 000020 000018 000008
1990 000011 000009 000004
1991 000023 000018 000009
1992 000011 000008 000004
1993 000018 000014 000006
1994 000012 000009 000004
1995 000010 000007 000003
1996 000009 000007 000003
1997 000018 000014 000006
1998 000010 000007 000003
1999 000013 000009 000004
2000 000005 000003 000001
2001 000015 000010 000005
2002 000013 000009 000004
2003 000010 000007 000003
2004 000017 000011 000005
2005 000020 000013 000006
2006 000011 000007 000003
2007 000012 000007 000003
Here are the numbers for seminomatous non-testicular tumors in white non-hispanic males.
crude of SEER of U.S. Pop
1973 000002 000003 000000
1974 000001 000001 000000
1975 000005 000006 000002
1976 000004 000005 000001
1977 000010 000013 000004
1978 000008 000010 000003
1979 000004 000004 000001
1980 000014 000016 000006
1981 000016 000018 000006
1982 000011 000012 000004
1983 000011 000012 000004
1984 000011 000011 000004
1985 000013 000013 000005
1986 000016 000015 000006
1987 000014 000013 000005
1988 000011 000010 000004
1989 000016 000014 000006
1990 000010 000008 000004
1991 000011 000008 000004
1992 000014 000010 000005
1993 000013 000010 000005
1994 000023 000018 000008
1995 000017 000013 000006
1996 000020 000016 000007
1997 000018 000014 000006
1998 000010 000007 000003
1999 000009 000006 000003
2000 000016 000012 000005
2001 000014 000010 000004
2002 000017 000012 000005
2003 000016 000011 000005
2004 000022 000015 000007
2005 000026 000017 000008
2006 000014 000009 000004
2007 000017 000011 000005
Non-testicular germ cell tumors represent a tiny fraction of the germ cell tumors occurring in men. For the purposes of analysis, there's not much you can do with these tumors. They're not going to give you statistically significant results when you try to test a hypothesis; some of them may represent misdiagnoses (e.g., colon cancer mistaken for monomorphic teratoma in a peri-testicular appendage), or a conservative topographic assignment (peri-testicular metastasis from a regressed primary germ cell tumor). Because we are trying to find a biological explanation for the rise in seminomas in white men, we can ignore the very small number of non-testicular germ cell tumors.
Definition problems
OK, the rate of occurrence of seminomatous germ cell tumors of the testes has been greatly increasing, in the white male population, since (at least) 1973. During the same period, the rate of occurrence of the other type of germ cell tumors (non-seminomatous) has hardly increased at all, for white men.
Why has the rate of occurrence of seminomas increased since 1973, in the white male population? Also, if seminomatous and non-seminomatous germ cell tumors are just morphologic variants of the same basic tumor (i.e., germ cell tumor), why wouldn't they both increase to the same extent?
Perhaps some of the problem relates to the definition of these two tumors. Seminomas are tumors of gonocytes, a differentiated cell committed to producing gametes (sperm in males, eggs in females), or a committed progenitor cell of gamete-producing cells (i.e., an ancestral cell of a gamete-producing cell). Since seminomas are considered the neoplastic equivalent of gonocytes, there seems to be little leeway in their classification: they must be included among the germ cell tumors.
But what about the other type of germ cell tumors. This other type is known by two different names that tell us a lot about the ambivalent nature of the tumor:
From wikipedia:
How can a germ cell tumor be non-germinomatous? Wouldn't the adjective "non-germinomatous" pretty much tell you that the tumor can't be a germ cell tumor?
It reminds me of one of my favorite limericks.
Well, what are the non-germinomatous germ cell tumors? These are tumors that usually arise in the gonads and are composed of primitive pluripotent cells. So we can find pure or mixed populations of embryonal carcinoma, teratomatous tissue, choriocarcinoma. These are the same cells that are found in the very earliest embryo and placenta. But these primitive cell types are not gonocytes (i.e., they are not differentiated cells committed to producing sperm or eggs). These tumor are primitive non-germ cells.
So why are the primitive non-germ cell tumors included among the germ cell tumors?
The answer to this question comes from our understanding of the common precancer of most of the seminomatous and non-seminomatous germ cell tumors: intratubular germ cell neoplasia.
The common precancer of testicular germ cell cancers: ITGCN
If you have a sub-class of germ cell tumors that are called "non-germinomatous germ cell tumors," isn't that a contradiction in terms? Isn't it like saying that dehydrated water is a subclass of water? There is a simple explanation: the classic germ cell tumors of the testes (seminoma), as well as most of the malignant non-germinomatous germ cell tumors of the testes, arise from the same precancer: intratubular germ cell neoplasia (ITGCN). Because ITGCN is composed of dysplastic (early neoplastic) germ cells, both the germinomatous and non-germinomatous tumors have a germ cell origin. You can easily appreciate the morphologic similarity between ITGCN and seminoma by looking at a histologic preparation of each.
The germ cell precancer, ITGCN, is a collection of atypical gonocytic cells lining seminiferous tubules in the testis.
Seminoma cells closely resemble the cells of ITGCN, from which they derive (with the rare exception of the so-called spermatocytic seminoma, which behaves unlike the other types of seminomas). The same precancer precedes the development of most of the invasive non-germinomatous germ cell tumors of the testis.
So, the terminologic mystery is solved. The germinomatous and the non-germinomatous germ cell tumors are classified together because most of them are derived from neoplastic intratubular germ cells (i.e., intratubular germ cell neoplasia).
But solving the terminologic mystery does not help us understand the biology of what's happening. Why does ITGCN give rise to tumors of germ cells (e.g., seminomas) and to tumors of primimitive non-germ cells (e.g., embryonal carcinoma, choriocarcinoma)? How can a tumor be derived from cells that have a committed lineage (i.e., sperm cells in the case of males) that is completely unrelated to the lineages found in the tumor?
There's an answer. It has a lot to do with a phenomenon unique to germ cells called epigenomic erasure.
As an organism develops, cells specialize into about 200 differentiated cell types. All these different types of cells have the same genetic sequence (genome). Cell types are distingued, one from the other, by epigenetic modifications. Epigenetic modifications to genes involve base methylation, conformational changes in chromosomes, protein modifications... anything other than changes in DNA sequence. Germ cells, like all other differentiated cells, have epigenetic modifications. The unique thing about germ cells is that they must undergo epigenetic erasure prior to the production of gametes; otherwise the gametes would be imprinted with the epigenetic modifications characteristic of the parent organism and would not be capable of recombining during fertilization to produce a fully de-differentiated, totipotent product.
The cells of intratubular germ cell neoplasia (the precancer of most male germ cell tumors) and of seminomas, are all characterized by DNA hypomethylation; not so for the cells of non-germinomatous germ cell tumors.[2,3] DNA Hypomethylation is seen in epigenomic erasure [of germ cells].
"Erased" germ cells are capable of developing into totipotent embryonic cells.[4] It would seem that a plausible mechanism for the development of non-germinomatous germ cell cancers from a germ cell precursor (intratubular germ cell neoplasia, ITGCN) is that the "erased" ITGCN cells, during cancer development, transform into totipotent cells, capable of differentiating into cells from any embryonic layer (e.g., embryonal carcinoma), or into extra-embryonic tissue (e.g., choriocarcinoma).
This explains why the ITGCN, the germ cell precancer, can give rise to both germinomatous (erased) and non-germinomatous (epigenetic-modified) cancers.
There is only one mystery left to solve (the original mystery that we started with in the beginning of this web page). If germinomatous and non-germinomatous germ cell cancers both arise from the same precursor, why is there a much greater increase in the rate of occurrence of seminomas compared with the rate of occurrence of non-germinomatous cancers, since 1973?
Here is a graph, produced from the SEER public use data sets, of the crude occurrences of seminoma and non-seminoma testicular germ cell tumors, in white males, since 1973.
The light blue bars are the seminomas, and the maroon bars are the non-seminomatous germ cell tumors of the testes. Since 1973, the seminomas increased from a number much lower than the occurrences of the non-seminomatous germ cell tumors; exceeding them in 1977. Since 1977, the crude occurrences of seminomas has greatly outpaced the occurrences of the non-seminomatous germ cell tumors of testes in white males.
Why? If both types of tumors are coming from the same precancer, why are their trends of occurrence non-parallel? Well, there are several possible answers. It is possible that some external influence has modified the step in the progression of precancer to cancer, to favor the occurrence of seminomas. However, it is also possible that their increases in occurrence are indeed parallel, and we're just not seeing it in our graph. Bray et al have looked at the incidence of testicular seminoma and non-seminoma germ cell tumors, by cohort (i.e., year of birth), not by year of occurrence.[1]
When the comparisons are based on cohort (comparing incidence for people born the same year), most of the differences vanish [between the incidence of seminomas and non-seminomatous germ cell tumors]. When do we see a birth corhort effect on tumor incidence? For cancers, a cohort effect is best observed when individuals born in one year are exposed (as a population group) to a causal agent that is different from the exposure of individuals born in other years. The cancers that result may occur at many different ages, thus erasing the cohort effect when the data is stratified by year of occurrence (as we had done the graph above). Only when you look at the birth cohort will you find a trend that may relate to a carcinogenic exposure.
OK. The birth cohort data reported by Bray et al would seem to indicate that some generational effect is acting on succeeding cohorts to produce a shared increase in the incidence of all testicular germ cell cancers. Furthermore, whatever is causing the generational effect is likely to be of short duration or differ significantly from year to year. Why is that? If the exposure of a carcinogen were of long duration or were the same from year to year, then every cohort would be exposed similarly, and there would be no birth-year specific effect.
So, now the mystery is: What are the conditions and carcinogens that might cause testicular germ cell tumors, which have changed, year-by-year, to produce the observed rise in these cancers in white males?
There is one class of conditions that is overwhelmingly associated with the development of germ cell tumors of the testis: disorders of sex development of the testes.[5] Among the conditions within this general group are testicular dysgenesis, testicular feminization (insensitivity to androgens), and cryptorchidism. Disorders of sex development of the testis raise the incidence of intratubular germ cell neoplasia or of gonadoblastoma, both of which are testicular precancers.
As you might expect, along with the observed increase in testicular germ cell cancers in white males, there has been an observed increase in the incidence of disorders of sex development in the same population. [5,6] These disorders are characterized by a retardation in the maturation of primordial germ cells, along with an apparent mitotic over-stimulation of these same cells: leading to a proliferative, precancerous condition.
Though there is no proof at the moment, we might expect that males who develop testicular germ cell cancer who have clinically normal testes, may harbor small foci of [clinically unobserved] germ cell proliferative lesions.
What has caused the increased incidence of disorders of sex development in the testes? We don't know, but we have a candidate: the ubiquitous plasticizer and endocrine disruptor, Bisphenol A. Bisphenol A is a synthetic estrogen used in the process of manufacturing plastics, and has been detected in the serum, milk, saliva, urine, and amniotic fluid of humans.[7] Because we get our daily dose of Bisphenol A from plastic bottles, one would expect that the levels of Bisphenol A in our blood would have increased steadily over the past several decades [coinciding with our increased dependence of plastic food and drink containers]. You might also expect that if Bisphenol A produced testicular cancers, you would see the largest increases in incidence among the wealthiest populations in the most industrialized nations [as we do].
Can we assume that Bisphenol A is causing the rise of incidence of testicular germ cell cancers? Absolutely not. All of the evidence, so far, is very weak (if it can be called evidence at all!). Still, nobody would suggest that Bisphenol A has much to recommend itself as a healthy addition to our diets. It seems prudent to try to limit our exposure to this compound when feasible, particularly among infants and pregnant women.
SUMMARY
Using the SEER public use data files, we have seen a large increase in the incidence of germ cell cancers of the testis in white non-Hispanic males since the first SEER observation year (1973) up to the most recent data year (2007). Along with the increase in seminomatous germ cell cancers was a lesser but parallel increase in the non-seminomatous germ cell cancers of the testis, when compared in birth cohort populations.
The seminomatous and non-seminomatous germ cell cancers, though derived from very different cell types (germ cells versus embryonic/extra-embryonic primitive cells) develop from the same precanceous lesion (usually intratubular germ cell neoplasia and sometimes gonadoblastoma). Precancerous germ cells are characterized by epigenomic erasure, and this "erased" state seems to allow precancerous germ cells to develop into seminomas or into tumors derived from totipotent stem cells.
Testicular precancers develop from disorders of sex development. The
incidence of disorders of sex development, like the incidence of
testicular germ cell cancers, has been rising. Though the cause for
the rise of disorders of sex development (and the concomitant rise in
testicular germ cell cancers) is unknown. However, the ubiquitous
appearance of the platicizer and endocrine disruptor, Bisphenol A, has
captured the interest of toxicologists and cancer researchers.
REFERENCES
_nodal_metastasis.jpg){kind=link}