How to induce cancer to someone, with current technology? The answer is so simple, that you will not like it, so please bear with me (because I will not limit myself to answer no. 1):

The answer is: make the target start smoking! 

Yep, that easy. And perfectly within reach.

Tobacco inoculates you with up to 70 carcinogens, being the no. 1 avoidable cause of cancer, inducing cancers from the lung, larynx, head and neck, bladder, kidney, esophagus, stomach and colon. However, as you know, even smokers haven't got as nearly as 100% rate of cancer. The percentage is something as 10-30%.

So this just goes to show you how hard it is to induce cancer the way you want. Because cancer is a disease that is caused by mutations on the DNA that will make the cells go haywire and proliferate uncontrollably. But even after the mutation, the cells have lots of mechanisms to heal that mutation, be it DNA repair systems, be it immune responses against the tumor cells.

And carcinogen induces DNA mutations, either directly or indirectly.

This means that a person that is not exposed to a carcinogen may develop cancer, and a person exposed to a carcinogen may not. It's a matter of probability, of likelihood. 

To increase the cancer rates to non-negliglible numbers, you would have to opt between two strategies:
a) Expose the target to a single, but super-high dose of the carcinogen (e.g. radiation from an atomic bomb)
b) Expose the target to low, but continuous / repeated, doses of the carcinogen

Both of these would increase the rate of detection.

So let's see some ways we could do this, with technologies available today... even though none will be as simple and accessible as tobacco.
**************************************************
**1) Exposure to chemical carcinogens**

This is the easiest way and it doesn't need much sophisticated technologies to achieve. 

For example, sneakly cover the air conditioning of the rooms where the target works with a layer of asbestos. This will increase the risk of lung and pleural cancer.

Or taint the meals of the target with known carcinogens. For instance, make them eat only cereals contaminated with the Aspergillus fungus, that produces aflatoxin, that increases the rate of liver cancer. 

Or make your own cocktail of chemical carcinogens and put them on the food or air supplies.

**Pros**: The access to these carcinogens may be extremely simple, depending on the carcinogen.

**Cons**: The efficacy will be extremely low. You would need to constantly taint the food and air of the target, which would increase the risk of detection.
********************************************
**2) Radiation**

This is your prefered method, but I don't think it is feasible. Contrary to what Pete has said, I have no knowledge of solid evidence demonstrating that constant exposure to CT scans (or to radiation produced by CT machines) increase the risk of cancer significantly... if it means doing a semestral or trimestral CT scan.

Even radiotherapy, with significantly higher radiation doses, will not increase this risk with great orders of magnitude, even though radiotherapy is more limited in time and is fine-tuned to reduce the risk of radiation-induced cancer to the max. 

But still, you would need very high doses of radiation **AND** you would need them to be focused **on a very concentrated and constant spot in the body**. Even today, radioncologists need to make "armors" and "masks" that completely imobilize the patient and need to constantly recallibrate the machine in order to be certain that the beam is always affecting the same area, or else the therapeutic benefit gets compromised.

There is just no way to achieve this on a moving target with current technology, without using such a dose of radiation that would affect all the neighborhood, increasing the chances of detection.
********************************************
**3) Radioisotopes**

But there is a way that you can make radiation find your target. Kingledion has beat me to the punch with that. Just make the person ingest substances containing radiation releasing isotopes. They will decay at regular intervals and innoculate a dose of radiation at each interval.

You can concentrate the radiation on a specific organ by binding the radioisotopes to molecules that concentrate on particular regions of the body. 

For example, the only place on our body that uses iodine is the thyroid. Doctors use 131 iodine as a means to concentrate radiation on the thyroid. 

Another example, doctors use radionuclides bound to a prostate specific antigen on diagnostic PET. This antigen concentrates the radiation on the prostate, making it easier to detect prostatic cancer by measuring the radiation released. 

Kingledion mentioned strontium 90, which is analogous to calcium, which would concentrate the radiation on the bone. This would be helpful, because the bone marrow (the place where you produce the blood) is a place with a high celular division rate and therefore, very sensitive to cancer induction (leukemias, etc...)

**Cons**: Radiation may be detected,as kingledion as mentioned.
*********************************
**4) Immunity**

It is true that immunosupression may lead to cancer, because then there will be no immune cells to detect cancer cells and destroy them. However, that level of immunosupression would be hard to induce without extremely high doses of myelotoxics or radiation. Furthermore, it would be easily detected before inducing cancer... because the target would contract many infectious diseases beforehand, that would make him/her see a doctor, which would easily detect the immunosupression with a blood test.

So I don’t think that’s feasible.
***********************************
**5) Epigenetic tampering**

We don’t solely depend on genetics. We also depend on epigenetics. Meaning, all the cells in our organism possess the totality of our genetic code. But which genes we express, on which organs and at which situations depend on epigenetics.

Epigenetics means that we may turn on or off a gene, by adding molecule radicals (namely methyl radicals) to the DNA sequence immediately preceding that gene.

So, if a cancer is promoted by a mutated gene, we could epigenetically turn it off. And vice-versa... we could epigenetically turn on genes that would increase the likelihood of cancer.

At this time, we don’t have a way to manipulate epigenetics at our will. Every anticancer epigenetic drug that I know of has failed on clinical trials. But it is not a stretch to believe that this technology will be available in a couple of years.

**Cons:** You would need to develop this drug on highly specialized laboratories. Either you make a drug that just tampers with **all** your epigenetic make-up (and so the consequences will be unpredictable)... or you develop a drug that epigenetically focuses just on one gene, but then you will have a problem. Because you will need to make that drug available to the target on a sufficient dose. Since you can’t make physical contact with him, it means you can’t innoculate the drug. So the drug should be able to traverse the digestive system without being digested and be absorbed on a sufficient dose to achieve the effect. I don’t think this is feasible for a stealthy assassination.
******************
**6)  Addiction**

Remember tobacco? Everyone knows that it causes cancer! Yet everyone still does it.

Why? Because amongst the carcinogens you have nicotine, which is also a neurotransmitter that causes a sense of well-being and therefore, addiction.

Processed meat is a carcinogen, but who can live without that yummy hamburger? You just need to have it. Just one won’t make a difference, right? Cancer just happens to the others, right? You can stop whenever you want, right?

Just make sure that you couple your carcinogen with an addictive substance or behavior. **Thereby you will ensure that the target will be exposed to the carcinogen on continuous and regular doses.** The target will **BEG** you for the carcinogen.

In fact, the target will make the job for you. They will rationalize everything, everything, to be able to continue their addiction.  This is truer if you make the addictive drug have a hangover effect, that will cause depression or fatigue, so the target will no longer be able to live without it.

This is easy. Every neurological substance that induces addiction may induce (sometimes fatal) abstinence syndromes.

Heck, at a certain time, you can even forget about stealth. The target will be crawling up to you. If he asks for anyone else's help, he/she will stop getting the drug. He/She will not have that, especially if he/she is a dictator that won't have anyone deny him/her something.

Just make your evil chemists develop a carcinogen and highly addictive neurotransmitter, like nicotine.
***************************************
**7)  Infection**

This is, according to my opinion, the best strategy.  Viruses may cause cancer because they are genetic invaders, that will cause DNA damage. Bacteria may cause cancer, because they may produce carcinogens internally.

Any microbiological laboratory would have the technology to genetically alter bacteria or viruses in order to turn them cancerigenous, if they would set their mind on it.

*Helicobacter pylori* is a bacterium that infects nearly half of the world’s population and it increases the risk of gastric cancer. We know there are strains that are more cancerigenous than others. Taint the food with a genetically altered *H. pylori* and you’re good to go.

Or why not alter a non-cancerigenous bacteria that is ubiquitous on our intestinal flora, like an *E. coli*?

Or create a hybrid flu virus. It would be airborne and extremely virulent. But give it genes that would incorporate on the host’s DNA in order to induce mutations. Or make it hijack the cellular machinery to produce and release carcinogens on circulation.

**Pros**: You would have to inoculate the target just once, thereby decreasing chances of detection. It could be airborne or foodborne (or, if you don't count that as physical contact, sexually transmited - just hire and inoculate a prostitute). Once inside the host, the microbe would self-perpetuate, without the need for more actions from the killer. You could just sit back and enjoy.

In the case of viruses, doctors wouldn’t be able to detect them, because they would be new viruses. They would only be found if the doctors knew *a priori* what viral antigens they should be looking for.

**Cons** Theoretically, a doctor could eventually detect the infection.

**Pro pro** But even if they did diagnose accurately those infections, that doesn’t mean they could treat them. Viral infections are extremely difficult to treat. And you could genetically engineer your bacteria with antibiotic-resistance plasmids.