Atif Khan, MD

Intraoperative radiation therapy (IORT) is a method of accelerated partial breast irradiation (APBI) that allows for treatment at the time of lumpectomy. Last fall, updated results from two Phase III studies – the TARGIT-A trial and the ELIOT trial – were published in The Lancet and The Lancet Oncology. Both trials, which compare IORT to standard whole breast irradiation, reported significantly higher recurrence rates in the IORT arms.

In addition to inferior local control, physicians often cite other limitations of IORT, including treating without complete pathologic information, unproven radiobiology, and a lack of image guided treatment planning.

Radiation oncologist Atif Khan, MD, of The Cancer Institute of New Jersey, New Brunswick, NJ, shares his views on IORT, the reasons behind its rapid adoption, and how the recently published data do not support a role for IORT in the management of early-stage breast cancer.

What are the types of IORT currently available?

Currently, there are two main types of IORT technologies on the market.  In one type, a low energy or “soft” x-ray source is used while another type uses high energy electrons. For both types, some type of applicator or cone is lowered into the surgical field to treat the lumpectomy cavities.

Despite the increase in local recurrences, the investigators continue to claim the validity of IORT. Do you agree with their conclusions?

To the credit of the IORT investigators, they have launched and completed huge randomized trials of whole breast irradiation vs. IORT. The early results from both trials, the TARGIT-A trial and ELIOT trial, clearly demonstrate a statistically significant inferiority for the IORT arms.  The concordance of these trials is itself remarkable. As with any trial, the data can be subjected to some very creative post-hoc interpretations. While it is true that the current margin of inferiority in the TARGIT-A trial is small and within the pre-specified threshold of inferiority, it is a statistical difference, and most worrying because it became apparent very early on—with a median follow-up of somewhere between two and three years.

Do the data raise concerns about the safety and efficacy of IORT?

Although these trials are often lumped together, there are differences in their design that are important to understand when it comes to safety and efficacy.  The TARGIT trial used the “soft” x-ray type of technology.  This x-ray type attenuates very rapidly in tissue, and really results in what some have called a “homeopathic dose” of radiation at a treatment depth of 1 cm, the minimum usual target depth of APBI. The ELIOT trial uses high-energy electrons to treat the approximated edges of the lumpectomy cavity.  They use 21 Gy of dose, which by any accounts is real dose.  In fact, based on recent radiobiological understanding of breast cancer, is probably too much dose.  So the TARGIT trial undertreats the breast volume at risk and the ELIOT trial overtreats it.

In addition, the TARGIT trial allowed for whole breast in the event that adverse risk features were found after pathology review; no such accommodation was allowed in the ELIOT trial. When we look at the results, the findings could have been anticipated. The TARGIT trial has an excess of local failures because it’s an undertreatment, and these predominated in the group of patients who were carefully selected for the trial and in whom no whole breast was felt to be necessary.  This is important because the use of whole breast therapy is a confounder in this study, so the fact that this very early and highly significant difference is being seen in women who were selected for treatment after careful pathological review is a strong signal of the inadequacy of this treatment in even these lower risk women.  Similarly, we also see higher rates of local failures in the ELIOT trial. They didn’t allow whole breast RT on this trial, so if you had positive-nodes or a positive margin, they just went with it to see what might happen.  In that way it was a more pure experiment. But it turns out, it’s probably important to look for the risk factors we have long-associated with local recurrence before triaging a patient for this treatment.  The ELIOT trial demonstrates the hazards of pre-pathology treatment triage in my mind.  Furthermore, they see rather high rates of fat necrosis, a marker of tissue injury in the breast.

Do you expect to see the difference in recurrence rates for WBI vs. IORT increase over time? Could the higher recurrence rates with IORT possibly translate to lower survival as time goes on?

I do expect the recurrence rates to diverge over time. However, at these low rates of recurrence overall, it is probably unlikely we will see lower survival. Still I don’t think a treatment option that results in sub-optimal breast control deserves a place at the table when we have so many good treatment options for these women.

How do you explain the rapid adoption of IORT given the higher recurrence rates being reported?

Often times, with respect to these new technologies, the marketing gets ahead of the science. Part of what I see out there is programmatic adoption of IORT as a way to market the institution and get patients in the door.  It’s regrettable but this happens in many settings.

What are the most concerning limitations of IORT?

We already discussed the huge issue of treating a woman with a certain type of radiation before we even review her pathology specimen and understand her risks. The other main issue is that these treatments are not guided by images.  In all the rest of radiation oncology, we always insist on knowing where the tumor is, where the critical organs are, and where the dose is going. That just doesn’t happen with IORT.  For example, if there is a gap or bubble between the applicator surface and the breast tissue, no one may see it, and a woman gets treated that way without a record of the treatment deficiency.  In contrast, with catheter-based brachytherapy those treatment images are reviewed during treatment planning and accounted for, and there will always be a permanent record of how the treatment was done. It’s a matter of quality assurance. By the way, ensuring conformance of the applicator within the breast tissue (i.e. this bubble issue) is a key thing we look for with catheter-based brachytherapy.

How does the radiobiology of IORT compare to catheter-based brachytherapy? How do you respond to claims that extrapolate data supporting brachytherapy to justify the use of IORT?

The radiobiology of brachytherapy is well described and validated.  The radiobiology of IORT does not conform to what we know about how cancer cells respond to radiation. These are very different treatments.  It is true that both only treat a part of the breast, but we have to very careful to not conflate these two approaches.

What role does IORT currently play in the treatment of early-stage breast cancer? What role do you think it will play in the future?

IORT is an important advance and should be explored as a modality of highly conformal radiation that can be delivered in the OR.  However, the current techniques and data do not support a role in early-stage breast cancer.  There may be other indications where this treatment may be beneficial, like certain abdominal malignancies.  Unfortunately, many of the current solutions are meant to only treat breast cancer patients, because that’s where “the market” is perceived to be.

Do you offer IORT as a treatment option for your breast cancer patients?  

We do not currently use IORT for breast cancer therapy, and have no plans to implement such a program.

Dr. Khan is a board-certified radiation oncologist at The Cancer Institute of New Jersey in New Brunswick, NJ. With assistance from his multidisciplinary teams, Dr. Khan aims to help patients understand their treatment options so they can make informed decisions.