4

Uncertainty

Fate shuffles the cards and we play.

ARTHUR SCHOPENHAUER

Fate is not an eagle, it creeps like a rat.

ELIZABETH BOWEN

Jason couldn’t remember a time when things had been okay at home. When he was a small boy, perhaps they had been. But then his father started to have violent mood swings; he would fly into a rage over nothing at all. There was never any violence, but Jason’s parents fought constantly, for years. Eventually, when Jason was 12, his mother packed up their things and left, taking him and his sister to live in a different state. Jason never saw his father again.

Years later, thinking back on that time, Jason would wonder if perhaps his father had already been showing the first signs of the sickness that would, eventually, claim his life.

When I met Jason, he was in his early 30s. He and his partner, Lauren, had been living together for two years. They were thinking of marriage, of having children, of a future spent together. But first, Jason wanted to find out if he had a future at all.

Jason showed me a letter he had received eight years earlier. Addressed ‘to whom it may concern’, the letter said that Jason’s father had been diagnosed with Huntington disease, that this meant that his relatives were at risk of also developing the condition, and that the recipient might want to consider speaking to a clinical genetics unit about this information.

Huntington disease23 is not rare by the standards of genetic conditions — it affects about 1 in 10,000 people. It’s a cruel condition, with a characteristic trio of effects: a movement disorder, psychiatric problems, and a decline of mental function. There is a great deal of variation in the way that HD symptoms start and progress, but a typical story is of someone who had previously been well reaching his or her early 40s and starting to show signs of being affected. At first, these are subtle, easy to dismiss as by-products of normal ageing and entry to the middle years of life, rather than something more sinister: clumsiness, apathy, anxiety. Other people may start to notice that the affected person is having involuntary movements — twitching of the limbs and face. Depression may set in; balance and the ability to do complex physical tasks are lost. Over time, the brain’s decline becomes worse and worse; eventually affected people lose the ability to care for themselves, to walk, to speak, and to swallow. All of this takes years; most survive a decade or even two after the first symptoms begin. It is a process of slow neurological demolition.

[1 As we shall see in chapter 7, it’s not at all unusual that George Huntington’s 1872 description of the condition that has borne his name ever since was not the first such description … or the second, or the third. Others had beaten him to it with publications in 1832, 1841 (and 1842), 1846, 1860, and 1863. Five others could have had their names attached to the condition, but didn’t. Three of them published their descriptions before Huntington was even born!]

From the earliest descriptions in the first half of the 19th century, the familial nature of the condition was recognised. HD is inherited in a dominant fashion: if you’re affected then each of your children have a 1 in 2 chance of inheriting the faulty gene and being at risk. There’s something special about the fault in the gene: HD is one of a class of conditions known as triplet repeat disorders. You’ll remember the genetic code, made up of groups of three base pairs of DNA. In Huntington disease, the problem lies with a repetitive stretch of DNA: CAG CAG CAG CAG CAG … This codes for the amino acid glutamine, so that in one stretch near the beginning of the protein you have glutamine-glutamine-glutamine-glutamine-glutamine … and so on. Most of us have 35 or fewer repeats — typically 15 to 20 — and no chance of developing Huntington disease. People with 36 to 39 repeats might develop HD, sometimes starting later than usual, and progressing more slowly. With luck, they might never be affected at all. People with 40 or more repeats will develop HD, if they don’t die of something else first.

There’s a twist to this tale. If the repeated section is 27 repeats or longer, it’s too much to cope with for the machinery in the cell that’s responsible for copying DNA while making an egg or sperm. Mistakes can be made. In this case, it’s a very specific type of mistake: the DNA copying mechanism has a risk of slipping — like a stuck record. CAG CAG CAG skip CAG CAG where was I? CAG CAG CAG … This means there’s a chance that the child will inherit a larger (or, sometimes, smaller) repeat than their parent.

For reasons we don’t understand, it makes a difference whether a man or a woman passes the expansion on. On the whole, it’s worse to inherit the HD repeat expansion from your father24 than from your mother. The chance of a reduction in size is much less, and the chance of an expansion much more. Since larger repeat sizes lead to earlier development of symptoms, one of the consequences of all of this is that HD can actually get worse through the generations in a family — with earlier onset and quicker progression. There are a number of conditions that do this; the phenomenon is called ‘anticipation’, despite the fact that early onset Huntington disease is hardly something to look forward to. It took a while to figure out this generational worsening, because the phenomenon is kind of weird and unexpected, and also because it’s quite easy for family trees to give the appearance of anticipation just by chance. This is in part because the first person in a family who gets the diagnosis of a rare inherited condition is often a child with severe problems; that child’s affected parent is probably more mildly affected. That looks like anticipation, and, although it isn’t, you may need to see a fair few families before you can rule the possibility out.

[2 You might think this is an example of the process for making an egg being superior to the process for making a sperm, but it seems like it’s not that simple. For some other triplet repeat disorders, such as fragile X syndrome and myotonic dystrophy, it’s the other way round — expansions mainly happen when the repeat is passed on by a woman.]

You can probably see where things are headed here. If anticipation in Huntington disease means earlier and earlier onset through the generations, eventually it might get to a point when it starts very early — and it can. ‘Juvenile’ HD starts before the age of 20,25 and children sometimes show symptoms when they are as young as five. The diagnosis of HD in such a young child is a double blow: almost certainly, his or her father has a smaller expansion, but one that means he, too, will develop the disease in time.

[3 People with juvenile HD have at least 50 repeats; the youngest affected children can have more than 60.]

One more thing before you put yourself in Jason’s shoes. For now, at least, Huntington disease has no cure. There are treatments that can lessen the symptoms, to be sure, but once the symptoms start, the condition is relentless, a slow-moving juggernaut that has only one destination.

Now … let’s say there is a 1 in 2 chance that this is your fate. We can do a test and give you an answer. Think you’d like to take the test?

For most people, it turns out, the answer is a very definite NO. Nearly nine out of ten people who know they are at risk choose not to find out. Their reasons vary. Many figure that, since there’s nothing they could do about the information that will change the outcome, there’s no real benefit to having the information. Some prefer uncertainty, and the possibility of a long and healthy life, over certainty that could be of a bad outcome. It’s not knowledge you can un-know once you have it.

This means that the people who come to see a geneticist about testing are something of a select group. Reading this, you may think that you know what you would do, and that you are part of that select group, but it’s quite possible that you are wrong. When people are asked whether, in principle, they would choose to be tested, about 80 per cent say yes — almost the reverse of what actually happens. This gap between what people say they would do and what they actually do when given a real-world choice has a name. It’s called the intention–behaviour gap, and it applies well beyond the deeply personal and life-changing decision to have testing for HD. If you’ve ever signed up for a gym membership and then found yourself not, in fact, going to the gym — there’s the gap.

Having said that, the group of people who do choose to be tested are generally very sure about that decision. Sometimes, like Jason, people wait for years before taking positive steps to have the test. But once you make it into my clinic room, it’s very likely you’ll go ahead — I can count on the fingers of one hand those I’ve seen who have reversed the decision at or after that point. When that does happen, it can be at the very last minute. Our department’s files hold a sealed envelope containing the HD result of a man who had the test done but, when he was called with the news that his result was available, changed his mind about receiving it. For more than 15 years, that envelope has sat waiting, unopened. He has probably lived out the answer — developing symptoms, or not — by now.

From time to time, I do see people who come for testing but don’t really want to know the answer. Mostly, they are in a situation that is similar to Jason’s: they are planning a family and don’t want to pass the condition on to their children. This opens up an unusual option: an exclusion test.

The gene we are interested in here, HTT, sits on chromosome 4. The idea of exclusion testing for HD is to identify if an embryo has inherited its copy of chromosome 4 from Jason’s mother, not his father. Remember that Jason has two copies of chromosome 4, one inherited from his father and one from his mother, and Jason’s father had HD, but his mother did not. Jason will pass on either his father’s chromosome 4 or his mother’s, to each of his children.26 If the embryo has its grandmother’s chromosome 4, it’s in the clear. The laboratory would make no effort to work out which of Jason’s father’s two copies of chromosome 4 harbours the faulty gene, because the information isn’t needed to achieve the desired result. That way, we could be certain the baby won’t have HD without revealing Jason’s status.

[4 It’s actually a little more complex than that, because of recombination — a process of creating a new, mix-and-match chromosome containing parts of each of a person’s two copies of every chromosome. So really we are looking to identify an embryo with the grandmaternal chunk of chromosome 4 that contains the HTT gene, not the whole thing.]

Jason and Lauren considered this option, but Jason had already decided that he needed to know what would happen to him before making plans for the future. When I met him, it was his second visit — he’d already met with Lisa Bristowe, the genetic counsellor with whom I work. They had talked through the issues, the reasons to be tested or not to be, the possible implications for insurance,27 and how Jason might deal with either of the possible results. Lisa had been sensitive to red flags: Is this someone for whom a bad result might be a truly devastating blow? Would he need extra psychological support around the time of the result? We had arranged for Jason to meet with a neurologist, who had found no signs of HD; this meant that, if the result were positive, it would not mean that Jason had HD, but that he was destined to develop it later. We’d offered an appointment with a psychologist as well, but Jason had declined this.

[5 As you can imagine, life-insurance companies are not big fans of Huntington disease. As I write, in Australia there is a moratorium on insurance companies discriminating against people on the basis of genetic test results, for life-insurance policies of up to $500,000 as well as some other types of insurance. This is voluntary, and we don’t know if it will be permanent.]

At the second meeting, we ran over some of the same ground, discussed medical issues, and arranged the test. Two tests, in fact: we always do predictive tests in duplicate, because of the consequences of getting things wrong, and because the biggest single cause of laboratory error is testing the wrong sample. Rare though it is for two individuals’ blood tubes to be mixed up, sending two separate samples, which are tested independently, makes it nearly impossible for this to happen.

Six weeks later, we met the couple again. Walking to the outpatients clinic that morning, I carried a sealed envelope containing Jason’s result; just before the appointment, when I knew that he had arrived at the hospital and was waiting to see us, I opened it.

I don’t believe in luck, not really. But Jason had come to see me part way through a freakish two year period during which nobody that I tested for HD received bad news — and he did not break the run.

For Lisa and me — although not, of course, for Jason and Lauren — this was a pretty routine, straightforward scenario. But the ability to do ‘predictive’ testing can throw up situations that are not straightforward at all. Consider identical twins who have a 1 in 2 risk of developing HD. One wishes to be tested, the other does not. By testing one, you have tested the other. We would not communicate the information to the other twin, of course … but what are the chances that she will inadvertently find out, or, to put it another way, how likely is it that the secret will be kept from her?

If, somehow, she is not told, how will she live with the knowledge that her twin knows the answer for both of them? Imagine being the twin who does not know, having an ordinary conversation with your twin and knowing all the time that she knows your shared fate — and could tell you the answer in a second, with as little as a nod or a shake of the head?

I’ve never encountered this situation, but from time to time we see people like Kim, a young man whose maternal grandmother had recently died of HD. His mother did not wish to be tested, but he did. Good news for him would tell us nothing about his mother28 — but if he had inherited the faulty gene, then so had she. In that case, she accepted his decision to be tested, and it turned out that his 1 in 4 chance (1 in 2 chance his mother had inherited the faulty gene, times 1 in 2 chance she had then passed it to him) did not come up, leaving his mother where she had started.

[6 Almost nothing. There’s a branch of mathematics called Bayesian probability, which is something of a favourite of geneticists. It allows us to combine different types of information to modify our assessment of the likelihood that something will happen. Without digging too much into the details, in this case, the finding that the man did not inherit an HD expansion shifts the probability that his mother had inherited an expansion from 1 in 2 to 1 in 3.]

What about testing children? It’s natural for parents to be deeply concerned about the possibility that their children may develop a condition such as HD in the future, and to have a desire to find out — driven mostly, perhaps, by the hope of receiving good news. Very soon after predictive testing became possible, the genetics community decided that we should say no to such requests. There are various reasons for this: concerns of genetic discrimination, and of stigmatisation; worries that children will be treated differently, in a way that will harm them. The most compelling reason, for me, is that by testing them we take away their option not to be tested. Knowing that most adults, given the choice, do not have testing, is it fair to a child to take that future possibility off the table?

All of this hinges on the fact that we have no treatments yet that change the outcome in HD, although there is a great deal of research directed at developing such a treatment. If we knew that there was a medication that could prevent the disease from taking hold, and that you needed to start taking it when you were a child in order for it to work, the rules would change immediately. There are other conditions in which the stakes, and, as a result, the rules, are different in this way. Take familial adenomatous polyposis (FAP), for instance. This is a condition in which hundreds to thousands of growths, called polyps, form in the colon. Left alone, colon cancer is inevitable; surgery to remove the colon is needed once the polyps are present. Screening with colonoscopies is needed to see if that has happened yet, starting from around 10–12 years of age. Polyps tend to start showing up in the mid-teens, but they can appear even earlier.

Because of this, genetic testing in a child who is at risk of having FAP is completely uncontroversial, although not taken lightly. There are important differences between FAP and HD. The age at which problems develop is generally much earlier in FAP — genetic testing in a child would not usually be happening decades before the first symptoms might appear. And you can do something about it, using screening and surgery … and the something that you can do is relatively burdensome,29 adding an incentive to do testing, so that you can spare half of the at-risk kids from having to go through it.

[7 I can attest from personal experience that having a colonoscopy is no big deal, thanks to the wonders of anaesthesia — but preparing for one (bowel washout) is not fun at all.]

Falling somewhere in between these extremes is a group of conditions that affect the heart. A cardiomyopathy is a disease of the heart muscle. Most commonly, the muscle becomes thickened,30 which can block the flow of blood through the heart; or it becomes weak and floppy.31 Either type can lead to problems with the flow of electricity through the heart, with potentially fatal results. Other conditions, such as long QT syndrome, affect that flow of electricity without changing the heart muscle.

[8 Hypertrophic cardiomyopathy.]

[9 Dilated cardiomyopathy. There are several other types of cardiomyopathy that are much less common than these two.]

These are mainly dominant conditions, like HD and FAP, and they vary enormously within and between families affected by them. I once saw a 14-year-old boy who popped into a fast-food restaurant on his way home from school. It happened that the woman standing in line behind him was a nurse. This was fortunate because, when the boy’s heart suddenly stopped, the nurse performed CPR until the ambulance arrived. Remarkably, the boy suffered no ill effects, but it turned out that he had quite a severe cardiomyopathy that had caused his cardiac arrest. We were able to identify the genetic cause of his condition and track it through the family. There were several people in the family who turned out to have relatively mild heart problems — including the boy’s mother. But her father, who was in his 70s, carried the same genetic change as his grandson, and had done so all his life, yet his heart remained healthy.

For adults, at least, deciding to be tested for conditions like this is generally much more straightforward than the decision to be tested for Huntington disease. The implications are worrying, to be sure, and it might be distressing to find out that you are at risk of developing a serious heart condition. But there are treatments that can change the likelihood that you’ll die from the condition, and there’s that chance that you might never develop symptoms. Still, though — should we test children in families like this, to see if they have inherited the at-risk version of the gene? What would the result mean? It’s not ‘predictive’ if you can live your whole life without ever developing a problem. On the other hand, these are not always adult-onset conditions — quite young children can be affected sometimes. There are treatments that can reduce the risks … but you could also just screen for problems using non-genetic tests such as echocardiography (an ultrasound of the heart), which are not invasive or burdensome, and then treat any problems as they arise.

It’s not at all clear what we should do about this,32 and different geneticists have come to different conclusions. My own position has shifted over the years; now, I explain all the issues to parents who want to have their children tested, I make sure they understand the implications, and then, if they still want to go ahead, I do the test. If the child is old enough, they get to have a voice in the decision-making process. Teenagers often say no.

[10 It’s not clear to geneticists. Cardiologists are generally very clear-cut about the issue: they want us to just get on with it, so that they don’t have to screen people who don’t need to be screened.]

The implications of the results are potentially a bit different for a heart condition than for a brain condition or cancer. Probably the risk of stigma isn’t there, but there are some extra concerns. Want to be an airline pilot when you grow up? How sympathetic is the company doctor likely to be to the idea that you have a genetic test result that says your heart might suddenly stop one day? Never mind that you might also live a long and healthy life and never show effects — aviation is a risk-averse business, and it wouldn’t be at all surprising if this type of information turned out to be career-limiting.

*

In HD, FAP, and inherited heart conditions, the uncertainty is often about whether to have a test. But sometimes, uncertainty follows from the result, rather than preceding the test.

Lee-Ann and Derek had been trying for a baby for what seemed like a long, long time. Test after test had not found a cause for their infertility, beyond one that had been obvious from the beginning: time was not on their side. When they first went to see a doctor about their difficulties conceiving, Lee-Ann was 37 and Derek was 40; by the time I met them, she was 41, and pregnant at last, a naturally conceived pregnancy after several years of unsuccessful IVF. They told me that they had been overjoyed by the pregnancy, but worried about the possibility of a chromosome problem in the baby. Then, an early ultrasound scan showed an excess of fluid at the back of the baby’s neck. Chromosomal conditions are among the possible causes for this finding,33 so they chose to have chorionic villus sampling (CVS) done. This test takes a tiny piece of the placenta for use in genetic testing, with the idea being that, since the placenta shares its genetic make-up with the baby, if the baby has a chromosomal problem, it will be present in the placenta as well.34

[11 There’s a long list of other possible causes, but if the chromosomes and 18-to-20-week ultrasound are normal, the outcome is usually a healthy baby.]

[12 Sometimes, there can be changes that are present in the placenta but not in the baby. More on this in chapter 11.]

Lee-Ann’s CVS showed that, instead of the usual 46 chromosomes, there were 45; instead of XX or XY, there was just a single X. This was the point at which I was asked to see them, to talk about what this result might mean for their baby.

In general — and there certainly are exceptions — if you have two X chromosomes, you’ll be a girl.35 In general — and there certainly are exceptions — if you have one X and one Y chromosome, you’ll be a boy. This is why the X and Y are called the ‘sex chromosomes’.

[13 Here — and elsewhere in the book — when I refer to girls and boys, and male and female, I am referring to physiological sex rather than gender identity.]

When its development begins, every embryo carries the potential to be either male or female. Early structures called the Wolffian duct, which has the ability to develop into male sex organs, and the Müllerian duct, which has the ability to develop into female sex organs, are present in every embryo. If, at around six weeks after conception, the SRY36 gene is activated, the Müllerian duct will wither and the Wolffian duct will develop. It’s a boy! As its name suggests, SRY is on the Y chromosome. If there is no SRY signal — for example, because there is no Y chromosome — a different set of signals kick in. The Wolffian duct withers, the Müllerian duct develops. It’s a girl!

[14 SRY, for ‘Sex-determining Region on the Y chromosome’.]

Except when it isn’t.

This is a complex process, and, like all complex processes, it has vulnerabilities — ways that things can go wrong. The ‘disorders of sexual differentiation’ (DSD)37 include a spectrum of variations on a theme, ranging all the way from a boy with two X chromosomes to a girl with an X and a Y. Infertility is often part of the picture, and sometimes there are other medical complications, because some of the genes involved are important not just for sex development but for other parts of the body as well. Sometimes in these conditions, the baby’s genitalia can be ‘ambiguous’ — when the midwife checks the baby, the best answer to the question ‘Is it a boy or a girl?’ may be ‘I don’t know’.

[15 Also known as ‘disorders of sex development’ and ‘differences of sex development’.]

Perhaps surprisingly, abnormalities involving the sex chromosomes are only rarely a cause of any doubt about the sex of the baby,38 although they can and do cause problems, in ways that relate to the special status of the X and Y. As discussed in chapter 1, for all of your chromosomes but that one pair, you really, really need to have two copies, and only two.

[16 The main scenario in which this sometimes happens is when the baby is a mosaic, with some cells having a Y and others not. Most commonly, this is a mixture of cells with 46 chromosomes, including X and Y, and other cells that have lost the Y and have 45 chromosomes with just the one X. Even in this situation, the most common outcome is a boy, although anything is possible — including a girl with Turner syndrome or a baby with ambiguous genitalia.]

But the X and Y are special. Or, arguably, the X is special and the Y just basks in its glory. The X chromosome is large, and packed with important genes, including many that are essential for the way the brain develops, and thus for intelligence. The Y chromosome is mostly junk. Its genes include SRY, a handful of genes needed to grow testicles and make sperm … and not much else.

Which leads to something of a mystery. Why is it okay for most men to have only a single copy of such a large, important chromosome — if having one copy of chromosome 21, a much smaller and less important chromosome, is uniformly lethal? Alternately, why is it okay for women to have two copies of the X chromosome if the single copy that men have is the correct number? Why isn’t the extra chromosome causing problems in women?

The answer to this question was proposed by Mary F. Lyon, a mouse geneticist. Lyon had completed her PhD in the 1940s in Cambridge, under the supervision of R.A. Fisher, working on genetic mapping in mice.39 In the years following World War II, she went on to study the chromosomes of mice that had been exposed to radiation, in work funded because of concerns about the possible effects of nuclear weapons on chromosomes. In a letter to Nature published in 1961, Lyon put together several previous observations, including coat colour patterns in mice with mutations that had been linked to the X chromosome,40 and the fact that mice with only a single X chromosome were apparently normal females. She deduced that, early in embryonic life, one of the two copies of the X chromosome in each of the cells of a female mammal is switched off — inactivated — and the other is active. If a female mouse has a variation in a coat-colour gene on one of its two X chromosomes, and the normal,41 or ‘wild type’, version of the gene on the other copy of the X chromosome, you can expect exactly what had already been observed in at least six different types of mice: a patchy pattern, with a mix of normal and mutant coat colours. Hair roots with the normal gene active will produce fur of one colour, and those with the mutant gene active will produce fur of a different colour.

[17 We’ll meet Fisher again in chapter 9. In an oral history interview recorded in 2004, Lyon made it clear that Fisher, a famous theoretician, was no great shakes as an experimental geneticist. All the same, supervising Mary Lyon’s PhD was arguably a major contribution to the field by Fisher, so his efforts in the lab were not wasted.]

[18 Just as there are tortoiseshell cats, there are tortoiseshell mice — and the reason, Lyonisation, is exactly the same for both species. There don’t seem to be any tortoiseshell humans, which is a pity.]

[19 Of course, you can’t really have an ‘abnormal’ coat colour, but the principle applies to X-linked genetic conditions.]

Lyon’s proposal became known as the Lyon hypothesis, and the process was named Lyonisation42 (although it is now more commonly called, prosaically, X-inactivation). Every single prediction Lyon made about the underlying biology has since been proved correct. Lyon recognised that her hypothesis might have implications for human genetics, but, at the time, there wasn’t much known about human X-linked conditions. Now we know about lots of them. While males are often severely affected by such conditions, sometimes there are no effects in females at all; or effects can be patchy, whether in skin or in another tissue. At one extreme are conditions that are so severe that very few or even no males are ever seen — you need to have one functioning copy of the X chromosome to survive.

[20 The word ‘lionised’, as in ‘lionised by the press’, isn’t so commonly used any more. But thanks to Mary Lyon, whenever I do come across it, I experience a moment of confusion, sometimes accompanied by an odd mental image of a tortoiseshell politician.]

There are two parts of the X chromosome that are not affected by X-inactivation — they are switched on in both copies of a woman’s X chromosome, and they have exact copies in the Y chromosome. These pseudo-autosomal regions (so called because they behave as if they are on one of the non-sex chromosomes, the autosomes) are at the tips of the chromosomes and are necessary for the sex chromosomes to behave normally in cell division during sperm formation in males. They aren’t terribly large — PAR1, on the tip of the short arm of the X and Y chromosomes, contains just 16 genes, and PAR2,43 at the other end of the chromosomes, contains just three. But that doesn’t mean they aren’t important.

[21 Some people reportedly have a PAR3 as well! But since most of us don’t, it probably isn’t very important.]

If the PARs were just padding at the tips of the chromosomes, it probably wouldn’t much matter how many X chromosomes you had, because the extra ones would be switched off. It also wouldn’t matter if you had just a single X chromosome. And it probably wouldn’t matter a great deal if you had multiple copies of the Y chromosome.

As it turns out, though, having an extra sex chromosome is not altogether harmless. Women with an extra X, so that they have 47 chromosomes with three Xs, tend to be tall for their families but otherwise are mostly just healthy people without particular medical problems. It’s entirely possible to live your whole life with XXX and never know about it, or need to know. On the other hand, it is pretty clear now that, compared with women with just two copies of the X, those with XXX have a tendency to learning difficulties, and may have behaviour problems in childhood, or even autism. Having an extra Y (XYY) seems to have quite similar effects: most such men lead normal lives and never find out about their extra chromosome. XXY is more likely to cause noticeable problems — boys and men with the associated condition, Klinefelter syndrome, don’t produce as much testosterone as usual, are more likely to struggle at school, and are infertile. Some of these problems are treatable with testosterone. Adding more chromosomes into the mix — 48,XXXY or 49,XXXXY for instance — does make things worse, as you might expect.

Even though it’s decades since these conditions were first described, we don’t have rock-solid information about all this, because of the way that most people with these differences in their chromosomes are identified in the first place. If you are doing fine, nobody is going to bother counting your chromosomes. This means that the people who have a chromosome test are not representative of the group as a whole — they are skewed towards the more severe end of a spectrum. This is known as ascertainment bias and is the bugbear of researchers studying any condition that can be mild or severe.

During the 1960s and 1970s, there were several studies of prison inmates that seemed to show that men with XYY were more likely to be incarcerated, and the idea that the extra chromosome made you more aggressive and thus more likely to be locked up persisted for decades. As recently as 2006, the question was still open enough that a group of Danish researchers conducted a large study that concluded that criminal convictions were indeed more likely in men with either XYY or XXY … but that this effect almost entirely went away when adjusted for poverty, which itself is linked to a higher likelihood of being convicted of a crime. Since they were studying only men who were known to have an extra sex chromosome, and since these people were tested for a reason, it seems very likely that ascertainment bias is enough to explain the increased chance of criminal conviction that still remained after adjusting for poverty.

There have been a couple of studies in which large numbers of babies were screened for extra chromosomes and those with sex chromosome differences were then followed over time — a heroic effort, considering how many years you need to wait before you really know how things have turned out. As you might expect, the problems seen in those groups are generally much milder than those seen in people who were diagnosed because a doctor thought they had a problem that might be caused by a chromosomal condition.

Lee-Ann and Derek didn’t need to know about the effects of an extra chromosome, because their baby was one short. In general, missing chromosome material is more of a problem than extra, and this situation was no exception. It’s thought that 99 per cent of all babies that are conceived with a single X chromosome are miscarried, often before a pregnancy is even recognised. A common problem in pregnancy is a build-up of fluid in the tissues of the developing fetus, and this can by itself be severe enough to cause a miscarriage. Girls with just one X chromosome who get to be born have a condition called Turner syndrome. The effects of this are very variable indeed. Girls with Turner are pretty much always short,44 and pretty much always infertile, but everything else about the condition is unpredictable.

[22 Growth hormone can help with this.]

Just how variable are we talking about? Well, a girl with Turner syndrome might be born with congenital heart disease and kidney malformations. Her neck might be webbed, and she may have a distinctive facial appearance. She might have significant difficulties at school — her intelligence would usually be normal, but there are some particular areas that she might struggle with, to the point of needing extra help. She might be shy, anxious, and reserved. Planning and decision-making might be particular weaknesses for her.

Or she might have only some of those problems; or she could be completely fine, apart from being short. She might grow up not suspecting that she had any kind of medical problem at all, and the diagnosis might be made only when she is having tests to find out why she is infertile.

When I met Lee-Ann and Derek, I explained all of this, including the statistics about the likelihood of different aspects of the condition affecting a baby diagnosed in this way, as well as the weaknesses in the way those statistics were collected. I also explained that the only parts of all this that we could give them extra confidence about were the heart and kidney abnormalities, which mostly should be detectable on ultrasound. The rest was unpredictable; they would be facing uncertainty, much of which would take years to resolve.

For Lee-Ann and Derek, the decision about what to do with this information was relatively simple. They told me that if the baby had had a lethal problem, they would have requested a termination of pregnancy, and that if the baby had had a condition like Down syndrome, it would have been a challenging decision for them, and they weren’t sure what they would have done. For them, though, Turner syndrome — even at the end of the spectrum with the most problems — seemed like a milder issue. They were sad about the prospect that their daughter would face extra challenges in her life, and the infertility that is part of the condition was particularly a blow to Lee-Ann. But they definitely wanted to continue with the pregnancy.

This decision is not as straightforward for most of the people I meet in similar situations. Many really struggle with the information, and, in the end, a majority of couples request termination of pregnancy after learning their baby has Turner or Klinefelter syndrome. Even when the finding is XXX or XYY, it’s not uncommon to choose termination.

In dealing with this kind of finding, the uncertainty about what will happen is often one of the hardest parts.

All of us have to make decisions in situations where the outcome is unknown — decisions about relationships, about education and careers. There’s a whole genre of self-help books about dealing with uncertainty, which speaks to how hard it can be. It’s true that every prospective parent faces uncertainties about what the child will be like, and how they will cope with parenthood. But few are presented with such a specific range of possibilities and asked to make a tough choice, knowing that either decision will have a huge impact on their lives. To make things harder, from the time of the result, there is pressure to decide soon, because, if you don’t, the pregnancy will be too far along and the choice will be out of your hands.

Uncertainty is a constant in clinical genetics, and it comes in various forms. Our stock in trade is situations like Jason’s, in which the possible outcomes are clear, and the uncertainty is about whether to have a test; or like Lee-Ann and Derek’s, in which the test has already been done, and the diagnosis is clear, but exactly what it will mean remains uncertain.

Increasingly, we find ourselves having to cope with another type of uncertainty. All too often, we do a test and find ourselves uncertain about whether the result means anything at all.