RAI Following Thyroid Cancer Surgery Increases Risk for AML & CML

In a recent study of the risk and outcomes of a second hematologic malignancy in a population-based cohort of patients with well-differentiated thyroid cancer (WDTC), researchers found an early increase in the risk of acute myeloid leukemia (AML) and chronic myeloid leukemia (CML), but no other malignancies, in patients treated with radioactive iodine (RAI) following surgery.

The increased risk of AML and CML was seen not only in patients with high-risk disease features, but also in those with low-risk and intermediate-risk tumors. The risk for AML and CML peaked during the first 2 years after RAI treatment. In addition, AML developing after RAI “trended toward inferior survival” compared to patients in whom AML was their first cancer (median overall survival, 1.2 years vs. 2.9 years; P=.06) (J Clin Oncol 2017; https://doi.org/10.1200/JCO.2017.75.0232).

Of 148,215 patients identified with WDTC from the 18 Surveillance, Epidemiology, and End Results (SEER) registries, 53 percent received surgery alone and 47 percent received RAI. In total, 783 patients developed a second hematological malignancy. Compared with thyroidectomy alone, RAI treatment was associated with an increased early risk of AML (HR, 1.79; 95% CI, 1.13 to 2.82; P=.01) and CML (HR, 3.44; 95% CI, 1.87 to 6.36; P=.001).

According to the researchers, patients were excluded from the study for the following:

  • their thyroid cancer was not follicular or papillary histology;
  • they received treatment with chemotherapy or tyrosine kinase inhibitors;
  • WDTC was not their first cancer;
  • their hematologic malignancy was a first, third, or higher order of primary cancer;
  • they received external beam radiotherapy; and
  • if radiation or survival status was unknown.

Competing risk regression analysis was used to calculate the risks of second hematologic malignancy treatment and outcomes in the study.

Because SEER does not report on the dose of RAI administered, it was not possible to directly determine dose-response relationship of radiation-induced leukemia. Consistent with RAI dosing recommendations that bases RAI dose on disease burden, investigators applied an interaction term based on tumor size and regional disease to assess dose-response. WDTC patients with tumor size ≥ 2 cm (vs. < 2 cm) and regional disease (vs. local) had higher risk of AML and CML, suggesting receipt of higher RAI dose.

Individualized Prognosis

In an interview with Oncology Times, the study’s corresponding author, Sudipto Mukherjee, MD, PhD, MPH, Associate Staff Member in the Department of Hematology and Medical Oncology at the Cleveland Clinic Taussig Cancer Institute, said the impetus for the study was to provide patients with WDTC a more complete picture of what to expect after diagnosis.

“Far too often in clinical practice we offer therapeutic options to cancer patients who need those treatments, but often fail to provide them with a comprehensive picture of benefits and risk of available therapies, particularly long-term risk of second cancers, which is a devastating complication of cancer treatment,” said Mukherjee, whose main focus of research is trying to quantify the risk for second cancers in patients who had their first cancer treated with chemotherapy, radiation, or both.

Several previous studies have examined the relationship between RAI and development of second cancers, both solid tumors as well as hematological malignancies. What separates this study from previous studies is that, in this study, the researchers quantified the risk of individual hematologic malignancies rather than using broad categories of leukemia and lymphoma and, secondly, showed dynamic changes in the risk of these cancers over time.

“Second cancer risk analysis done by grouping several different leukemic (acute and chronic) entities under the broad category of leukemia is an oversimplified analysis and findings using such an approach can be potentially misleading. This is because it does not account for the fact that leukemias are biologically heterogeneous with different prognosis and outcomes,” Mukherjee said.

“For example, if a patient presents with AML after RAI treatment, providing them with an AML-specific prognosis is more accurate rather than overall acute leukemia risk that groups AML as well as acute lymphoblastic leukemia (ALL). AML and ALL, as we know, are disparate leukemic entities with different treatment strategies and outcomes—they are apples and oranges,” he explained.

The study also reports on dynamic changes in the risk of developing a second hematologic malignancy by comparing it with the risk of such cancers in the general population, Mukherjee noted. “I think a much more meaningful way of explaining the risk to the patients is, ‘Well, you have an increased risk of developing leukemia following the treatment of thyroid cancer, your risk of leukemia peaks in the second year of completing your treatment and, after 5-6 years following RAI exposure, your risk drops to baseline population rates.’ Providing a sense of how the risk changes over time following treatment gives patients a better understanding of what to expect with increasing number of years survived.”

Clinical Implications

The study’s results support the most recent American Thyroid Association guidelines, which instruct physicians to use caution when treating WDTC patients with low- or intermediate-risk tumors with RAI. There were two surprising findings, however, Mukherjee said.

“One, I was not expecting the risk of AML would peak so early. We found that AML risk peaks after the second year following exposure to RAI, which is a very early rise, and by 6 years the risk declines and becomes comparable to what we would see in the general population.”

A second surprising finding was an extended risk for CML, he said. “We found that the risk for CML peaked in the second year, but the elevated risk persisted for up to 10 years [before] it starts declining, which is something we did not expect.”

A major finding of clinical concern in this study was increased risk of AML and CML with RAI treatment even in low- and intermediate-risk WDTC, which comprise 94 percent of all WDTC tumors, Mukherjee said.

“If you look at our study, which encapsulates 40 years, from 1973 to 2014, the use of RAI following surgery for WDTCs went up steadily from 3 percent to 4 percent in 1973 to 50 percent in 2006. A substantial majority of these tumors are small (< 2 cm), asymptomatic tumors classified as low-risk for which use of RAI following surgery has not been shown to improve survival. But, definitely, these patients are put at risk of having a second cancer, particularly AML and CML.”

Fortunately, the ATA guidelines calling for caution in using RAI in low-risk tumors appear to be increasingly recognized in real-world clinical practice. “Since the issuance of these guidelines, we have started to notice a decline in RAI use for these tumors, which dropped to about 46 percent by 2014,” Mukherjee said.

“The clinical implications of these findings really are that, for WDTC patients with high disease risk features, RAI is necessary because of survival benefit with this approach,” he added. “But our study clearly shows that the risk of AML and CML following RAI also increases in low- and intermediate-risk WDTC tumors treated with RAI. I think that is the key take-home message—that, if it is possible, RAI should be avoided in low- to intermediate-risk tumors, in light of data that has not shown any survival benefit in these disease categories to date.”

The second take-home message is that patients with WDTC exposed to RAI should be actively monitored for myeloid malignancies, particularly AML and CML, during their cancer surveillance, Mukherjee continued. “We did not delve into that, but it seems like regular monitoring of the peripheral blood counts is one of the simplest ways to monitor these patients, especially in those patients who do not recover their counts completely in a reasonable period of time following RAI treatment, at least for the first 2 years, because the risk peaks during the second year.”

Further Studies Needed

Although the number of patients with WDTC who develop second cancers following RAI treatment is small, and even though the absolute risk for developing AML and CML is low, the problem is real and potentially avoidable. Especially for AML, this is concerning as the prognosis is dismal, Mukherjee said.

“What I would be curious to know is: Are there any other factors in addition to having RAI that puts this select group of patients at risk of having AML?” he said. ”Do these patients who develop these cancers following RAI exposure have any kind of underlying genetic predisposition to develop these cancers? This is an area that needs further investigation, because now we have data from other diseases that supports this assertion.

“For example, several recent publications have shown that people with CHIP [clonal hematopoiesis of indeterminate potential] mutations are at increased risk of developing myeloid malignancies,” said Mukherjee, pointing to two studies published last year that showed patients with solid tumors who harbored CHIP mutations at the time of their diagnosis were at much higher risk of subsequently developing therapy-related myeloid malignancies (Lancet Oncol 2017;18(1):100-111, Lancet Oncol 2017;18(1):112-121). “This is something that has not really been explored in the thyroid area.”

Meanwhile, potential alternatives to RAI are on the horizon. “I think that a lot of these options will be targeted therapies driven by genomic discoveries. This together with increasing awareness of the risks of RAI will hopefully lead to decline in the use of RAI for low-risk WDTC tumors over the next decade with a concomitant decline in the risk of fatal second cancers like AML in long-term WDTC survivors.”


Melanie Goldfarb, MD, Endocrine Surgeon and Director of the Endocrine Tumor Program at John Wayne Cancer Institute at Providence Saint John’s Health Center in Santa Monica, Calif., said the results of the study by Mukherjee, et al, re-enforce the position of many high-volume endocrine surgeons like herself who are confident they will not injure the recurrent laryngeal nerve, which controls the vocal cords, or harm the parathyroid gland, and thus are able to only use RAI in patients with a high risk of disease recurrence.

“There was definitely a time when every patient got RAI following surgery, but that has tapered off,” Goldfarb told Oncology Times. “If the thyroid cancer is growing into other tissues in the neck or into other parts of the body, if it is growing outside of the lymph nodes, or if they have really ugly histology, that’s really the only time we are going to say, ‘Hey, we should really use some RAI.”

“This is an important retrospective population-based study which adds growing evidence that adjuvant RAI therapy is a risk factor for therapy-associated AML and, interestingly, CML,” said Jack Jacoub, MD, Medical Oncologist, Hematologist, and Medical Director at MemorialCare Cancer Institute, Orange Coast Medical Center, Fountain Valley, Calif.

“[CML] is associated with a specific acquired genetic abnormality, and the association with environmental risk factors such as radiation is much weaker than for AML. Nonetheless, these findings should give pause to neck surgeons, nuclear medicine doctors, oncologists, and endocrinologists when evaluating patients with small [WDTC],” he said.

Chuck Holt is a contributing writer.

Leave a Reply

Your email address will not be published. Required fields are marked *