Guidelines, predictive models needed to improve understanding of second primary malignancies

Cancer survival rates are at an all-time high due to improved detection, greater understanding of the genetic drivers of diseases, and the development of more effective, targeted treatments.

The number of cancer survivors in the United States, estimated at 13.7 million in 2012, is expected to reach 18 million — a 31% increase — by 2022. By that time, about two-thirds of cancer survivors are going to be aged 65 years or older, according to the American Association for Cancer Research’s most recent Annual Report on Cancer Survivorship in the United States.

The trend has triggered a seismic shift in the landscape of cancer care, as the clinical community — bracing for a considerable shortage of oncologists — searches for innovative ways to care for patients who have completed treatment but still have specific medical, psychological and social needs.

A direct consequence of extended survivorship — the development of second primary malignancies — has the potential to further burden the health care system.

An estimated 1.6 million new cancer cases are diagnosed each year, and 16% of them are second cancers, according to a SEER report. Of those second cancers, more than 80% arose in separate or independent organ systems. The report also concluded that cancer survivors have a 14% higher risk for developing a new malignancy than the general population.

Experts cite several potential explanations:

• Genetic mutations make certain individuals more susceptible to malignancy;
• Toxic treatment for the first cancer contributes to the development of a second, either through radiation exposure, immunosuppression or other factors;
• Increased surveillance of patients who complete cancer treatment translates to higher rates of detection; and
• Because survivorship is on the rise, those treated for cancer are living long enough for a second malignancy to develop.

“While a certain fraction of the subsequent tumors would be expected to develop at the same rate as in the general population, the patterns of excess risk that emerged are sufficiently distinctive to suggest risk factors that may be shared by the primary and subsequent tumors, or an effect of cancer therapies that are potentially carcinogenic,” Joseph F. Fraumeni Jr., MD, founding director of the NCI’s Division of Cancer Epidemiology and Genetics, and colleagues wrote in the SEER report.

Extensive research has been conducted on second primary malignancies, but no data set has provided the clinical community with a comprehensive overview of the topic. Attempts are underway — primarily through the Childhood Cancer Survivor Study — to track lifetime risks among tens of thousands of individuals who may be likely to develop a second malignancy, but it could be years — or even decades — before answers emerge.

“A lot of small studies have been published in this area,” Matt Kalaycio, MD, FACP, chair of the department of hematologic oncology and blood disorders at Cleveland Clinic, told HemOnc Today. “But there are no landmark papers guiding this field.”

HemOnc Today spoke with a cross-section of clinicians and researchers to determine the potential role for each of these factors, and how they may contribute — both independently and in conjunction with each other — to the rise in second cancers.

‘Genes don’t change’

Certain genetic patterns that increase risk for secondary malignancy have been well described.

Malone and colleagues conducted a nested case-control study that involved more than 2,000 women with breast cancer. The results, published in 2010 in the Journal of Clinical Oncology, showed risk for contralateral breast cancer was 4.5 times higher (95% CI, 2.8-7.1) among BRCA1 mutation carriers and 3.4 times higher (95% CI, 2-5.8) among BRCA2 mutation carriers.

“Genes don’t change,” Axel Grothey, MD, of the division of medical oncology at Mayo Clinic in Rochester, Minn., said in an interview. “Carriers of the BRCA mutations have a higher risk of breast or ovarian cancers. Their lifetime propensity toward those malignancies is higher, and that includes a primary breast cancer or contralateral disease.”

The risk also extends to new primary malignancies.

“Inherited syndromes like the one caused by BRCA put individuals at risk for more than one kind of cancer,” Kalaycio said. “They can cause multiple cancers in various organs over time, irrespective of any therapy administered or other risk factors.”

Germline mutations in p53 increase risk for subsequent cancers, according to Leslie L. Robison, PhD, chair of epidemiology and cancer control at St. Jude Children’s Research Hospital.

“In children, genetic predisposition, alone or in combination with treatment-related exposures will be important in determining subsequent risk,” Robison told HemOnc Today, noting that survivors of bilateral retinoblastoma demonstrate one of the highest rates for developing subsequent neoplasms. “The germline p53 mutation in Li-Fraumeni Syndrome is hallmarked by pedigrees with individuals with soft tissue sarcoma during childhood.”

Besides its relevance during treatment, genetic information also can guide decisions about which cancer survivors should undergo more intense screening for new malignancies.

“If a patient has a familial gene that puts them at risk for malignancy, this should impact our risk assessment of the primary tumor and our subsequent screening approaches,” Grothey said. “If a 20-year-old gets colorectal cancer, they have an 89% risk of a later cancer. This makes a strong case for more screening, and for conducting a colorectomy rather than just removing the tumor.”

Increased surveillance

Because cancer survivors tend to undergo more frequent and intense medical follow-up, increased surveillance is another logical explanation for detection of secondary malignancies in this patient population.

“We are finding more malignancies, either intentionally or unintentionally,” Alfred I. Neugut, MD, PhD, Myron M. Studner professor of cancer research at the Herbert Irving Comprehensive Cancer Center at Columbia University, said in an interview. “We are doing many more CT scans for various reasons, and we find a lot of incidental cancers with all of the testing we do.”

Common examples of secondary cancers detected this way include prostate cancers and thyroid cancers, according to Luc G.T. Morris, MD, MSc, assistant attending physician at Memorial Sloan-Kettering Cancer Center.

“We occasionally pick up some other cancers that are quite indolent and may have otherwise gone unnoticed and never caused any problems for that patient,” Morris told HemOnc Today. “Most likely, the reason for the increased numbers of prostate or thyroid cancers is not that the other cancer caused prostate or thyroid cancer, but simply that we diagnosed them in the cancer patients and otherwise would not have.”

Neugut framed the issue in more practical terms.

“Consider the individual who will develop lung cancer at age 75 and die at 76,” he said. “If we give that person a PSA at age 60 and we find a prostate cancer, we treat it. Now when the person develops lung cancer — the cancer from which he or she is actually going to die — it’s their second cancer.”

Although the potential for overdiagnosis remains a legitimate concern, evidence highlights the value of targeted screening and surveillance for cancer survivors.

Park and colleagues reviewed the National Health Insurance Corporation Study cohort and identified 14,181 men with a first cancer. Of these men, 204 developed second primary malignancies. The researchers evaluated smoking history, obesity and insulin resistance in this cohort.

Results showed smoking history was associated with increased incidence of second primary lung cancer (RR=3.69; 95% CI, 1.35-10.09). Obesity was associated with increased risk for second primary colorectal cancers (RR=3.45; 95% CI, 1.5-7.93) and genitourinary cancers (RR=3.61; 95% CI, 1.36-9.54). Elevated fasting serum glucose concentration was associated with increased incidence of second primary hepatopancreatobiliary malignancies (RR=3.33; 95% CI, 1.33-8.37) and smoking-related malignancies (RR=1.93; 95% CI, 1.01-3.68).

The increased availability of effective treatments also is a factor, as higher percentages of patients are surviving their first cancers.

“Ten years ago, a patient with multiple myeloma wouldn’t have been recommended for a mammogram because it was not expected that she would live long enough for a second malignancy to develop,” Kalaycio said. “Now, multiple myeloma is a chronic disease. These patients are receiving mammograms and we are finding second primary cancers.”

The relationships between certain types of malignancies also must be considered, Neugut said.

For example, among survivors of head and neck squamous cell carcinoma, second primary disease is the leading cause of death.

“We know HNSCC survivors are at dramatically increased risk of developing second head and neck cancers, or cancers in the lung or esophagus, and this affects how we manage surveillance for second cancers in these patients,” Morris said.

Data from the National Lung Cancer Screening Trial suggest that patients with head and neck cancers should be screened for lung cancers.

“Our group also found that HNSCC patients are at increased risk of developing colon cancer, but the absolute number of cases was so low that it would be unlikely to alter any screening guidelines for colon cancer, so this is not always an easy question to answer,” Morris said.

Treatment effects

Most clinicians recognize that cancer treatments have been associated with carcinogenesis.

“Through follow-up of survivors of childhood cancer, we have a good understanding of the major impact of radiation chemotherapy,” Robison said. “These are important risk factors for subsequent malignancies.”

Nottage and colleagues conducted a nested case-control study in which they reviewed data from 13,048 patients treated for pediatric cancer at St. Jude Children’s Research Hospital.

The researchers identified 19 cases of colon or rectum adenocarcinoma from that cohort. The investigators matched those patients with two control groups. The first included 148 controls for age at primary malignancy and follow-up interval, and the second included 72 controls matched for primary diagnosis in addition to the same criteria established for the first control group.

Nottage and colleagues used exact conditional logistic regression to calculate ORs for chemotherapy and radiation exposure.

The results, published in 2012 in the Journal of Clinical Oncology, showed 40-year cumulative incidence of secondary colorectal cancer was 1.4%, translating to a standardized incidence ratio of 10.9 (95% CI, 6.6-17) vs. the general population. Researchers also determined secondary colorectal cancer occurred more frequently in an irradiated segment of the colon (OR=7.7 for group 1, P=.001; OR=15.2 for group 2, P=.002). Each 10-Gy increase in radiation dose was associated with a 70% increase in risk. Exposure to alkylating agents was associated with an 8.8-fold increased risk for secondary colorectal cancer, and increased radiation volume also was linked to increased risk (OR=1.5 for group 1, P<.001; OR=1.8 for group 2, P<.001).

The issue is not whether radiation causes cancer, but rather how that affects clinical decisions and whether risk can be mitigated.

“We have to treat patients with the best therapy available,” Kalaycio said. “Radiation exposure increases the risk of malignancy. The exposure doesn’t even have to be the whole body. If you radiate your hand, those cells circulate through the body. If they mutate, you are at risk. But there is a dose relationship. We need to be mindful of how much radiation we give. If there is any way to keep the dose low, we should.”

Morris described radiation treatment as a continual “balancing of risks and benefits,” such as the practice of not using radioactive iodine in patients with very low-risk thyroid cancer.

“Understanding second primary malignancies also has to inform our use of cancer therapies that increase the risk of second primary malignancy,” Morris said.

Problems arise, though, with each subsequent malignancy due to the location of and cumulative effect of radiation exposure — a reality several clinicians referred to as a “catch-22.”

“Radiation to the pelvis is associated with higher risk of prostate or rectal cancers,” Grothey said. “Those risks are connected. If you have had prior radiation for prostate cancer, you can no longer have radiation for rectal cancer. We can’t deliver that standard of care for the second malignancy. A second treatment is eliminated.”

The clinical and research communities should focus on developing and treating patients with therapies that are less oncogenic or leukemogenic, Kalaycio said.

“The future may lie with targeted therapies like imatinib (Gleevec, Novartis) or ibrutinib (Imbruvica, Pharmacyclics) that don’t lead to DNA damage,” he said. “These therapies work on biologic pathways. We are seeing encouraging data with regard to short-term remission rates, and this may be beneficial in the long term because they do not seem to be associated with genetic mutation.”

Aging population

Age also is a risk factor for secondary primary malignancy.

Incidence of second primary disease has increased about 5% or 6% in the past few decades, Neugut said.

“The entire population is older,” Neugut said. “If the population lives to age 60, there will not be as many second cancers. If 40% of people get cancer at age 80, as life expectancy increases to 90 or 100 years, the number of secondary malignancies will increase, as well. This is an age-related phenomenon that has nothing to do with treatment whatsoever.”

Grothey agreed.

“Age is the most important risk factor for some cancers, including those in the colon and prostate,” he said. “It is inevitable, and not anything we could eliminate.”

Morris, however, remains unconvinced.

“We cannot attribute second primary malignancies to patients living longer,” he said. “The reason is that we measure second primary malignancies as a ratio of observed-to-expected cases. When we say some group of cancer survivors is at an increased risk of getting a second cancer, we mean they are at an increased risk compared with people of similar age, gender, race and time period. We know that there is a certain rate of cancers developing in the population, so when we study [second primary malignancies], we study the rate of second cancers in excess of the background rate.”

Effective management of childhood malignancies also contributes to age as a risk factor for a second primary malignancy, Robison said.

“If a patient is cured of a childhood cancer and goes on to live another 60 or 70 years, there is a substantial interval to manifest a subsequent primary,” he said. “Whether the risk for a second malignancy is related to the underlying cause of the original cancer or due to cancer treatment, or a combination of both, is difficult to predict. From an epidemiological standpoint, we will be able to begin to distinguish which risks are most associated with the first cancer and its treatment, and what might be attributed just to longevity.”

Extensive research is underway to attempt to describe second malignancy risk in childhood cancer survivors, Robison said.

“We are looking at tens of thousands of patients, and our results have characterized quite nicely the risks and what cancers are likely in the first 2 to 3 decades after a first malignancy,” he said. “We just have to wait to see what those risks will be within the fourth, fifth and sixth decades.”

Lifestyle choices

One variable largely outside clinicians’ influence that affects risk for secondary primary cancers is patient lifestyle.

“We are happy to send a patient off after they are cured, but there is not necessarily a routinely established time for discussion of lifestyle choices,” Grothey said.

Li and colleagues conducted a population-based nested case-control study to assess the associations between several lifestyle choices and risk for second primary invasive contralateral breast cancer among breast cancer survivors.

The results, published in the Journal of Clinical Oncology, showed current smoking (OR=2.2; 95% CI, 1.2-4), consumption of seven or more alcoholic beverages per week (OR=1.9; 95% CI, 1.1-3.2) and obesity (OR=1.4; 95% CI, 1-2.1) were associated with increased risk.

The findings make sense, Robison said.

“Tobacco can cause multiple different kinds of cancers,” Robison said. “If the first cancer is driven by a specific etiology, the second one may likely be from the same etiology.”

Morris agreed, citing the concept of “field cancerization,” a term coined by Danely P. Slaughter, MD, and colleagues in 1953.

“The classic example of this is HNSCC,” Morris said. “These patients get HNSCC from using tobacco and alcohol. Using tobacco and alcohol exposes the entire upper aerodigestive tract to the same carcinogens; therefore, patients with cancers in any of these areas are more likely to develop cancers in other areas. So HNSCC survivors are at increased risk of getting another separate head and neck cancer, or a lung or esophageal cancer, and vice versa.”

Ideally, providers should strive to educate patients about what behaviors can be safely re-established after treatment for a first cancer and which should be avoided, Grothey said.

“We can’t change our genes, but we can change our lifestyle,” Grothey said. “We all want to have these conversations with patients, but we are not always successful. More dedicated guidelines with regard to this would be helpful.”

Moving forward

Robison suggested the development of a risk prediction model would be an important step forward in targeting populations for screening and early detection.

Kovalchick and colleagues used data from the Childhood Cancer Survivor Study and two nested case-control studies to develop three absolute risk models for second primary thyroid cancer to help clinicians manage survivors of childhood malignancy.

“Model M1 included self-reported risk factors, model M2 added basic radiation and chemotherapy treatment information abstracted from medical records, and model M3 refined M2 by incorporating reconstructed radiation absorbed dose to the thyroid,” Kovalchick and colleagues wrote in the Journal of Clinical Oncology.

They concluded that model M2, with basic prior treatment information, may be a valuable tool to monitor thyroid cancer risk among childhood cancer survivors.

More research of that type is necessary, clinicians said.

“The most informative approach is to identify patients early for more aggressive screening to get them into programs designed to detect their disease earlier, at a more treatable stage,” Robison said. “We need to develop a comprehensive series of parameters that predict the likelihood of a second cancer so clinicians can make an evaluation on a patient-by-patient basis and make recommendations accordingly. The level of information for childhood cancer survivors is quite robust. We need to translate this into guidelines and predictive models.” – by Rob Volansky

References:

de Moor JS. Cancer Epidemiol Biomarkers Prev. 2013;22:561-570.

Kaufman EI. J Clin Oncol. 2008;26:392-398.

Kovalchick SA. J Clin Oncol. 2013;31:119-127.

Li CI. J Clin Oncol. 2009;27:5312-5318.

Malone KE. J Clin Oncol. 2010;28:2404-2410.

NCI. New Malignancies Among Cancer Survivors: SEER Cancer Registries, 1973-2000. Available at: http://seer.cancer.gov/archive/publications/mpmono/MPMonograph_complete.pdf. Accessed on March 24, 2014.

Nottage K. J Clin Oncol. 2012;30:2552-2558.

Park SM. J Clin Oncol. 2007;25:4835-4843.

For more information:

Axel Grothey, MD, can be reached at Division of Medical Oncology, Mayo Clinic, 200 First St. SW, Rochester, MN 55905; email: grothey.axel@mayo.edu.

Matt Kalaycio, MD, FACP, can be reached at Cleveland Clinic Main Campus, Mail Code R32, 9500 Euclid Ave., Cleveland, OH 44195; email: kalaycm@ccf.org.

Luc G.T. Morris, MD, MSc, can be reached at Memorial Sloan-Kettering Cancer Center, 1275 York Ave., New York, NY 10065; email: morrisl@mskcc.org.

Alfred I. Neugut, MD, PhD, can be reached at Columbia University Mailman School of Public Health, 722 W. 168th St., Room 725, New York, NY 10032; email: ain1@columbia.edu.

Leslie L. Robison, PhD, can be reached at Epidemiology and Cancer Control, MS 735, Room S6010, St. Jude Children’s Research Hospital, 262 Danny Thomas Place, Memphis, TN 38105-3678; email: les.robison@stjude.org.

Disclosure: Grothey, Kalaycio, Morris, Neugut and Robison report no relevant financial disclosures.Disclosure: van Londen reports no relevant financial disclosures.

 

 

Do screening guidelines go far enough to ensure survivors of childhood cancer are monitored for second malignancies?

In a few malignancies, there is compelling evidence to support initiation of surveillance for adult-onset cancers at younger ages than would be recommended for the general population.

For some histologic subtypes, we have had the opportunity to characterize the natural history of when that risk becomes elevated, as well as the clinical and treatment factors that contribute to that risk.

For example, young women treated with moderate- to high-dose chest radiation for childhood cancer have an incidence of breast cancer that is comparable to that of women who are BRCA mutation carriers. Radiation dose and breast volume in the radiation field contribute to this risk.

Research has shown that breast cancer risk becomes elevated about 8 years after radiation exposure — long before the age at which population breast cancer screening is recommended. Currently, evidence is insufficient to demonstrate that early initiation of breast cancer surveillance in these women reduces breast cancer mortality, but this surveillance may facilitate detection of small and early-stage tumors. Early detection may improve prognosis in some women with more limited treatment options because of prior exposure to radiation or anthracyclines.

This information has directly informed consensus-based clinical practice guidelines developed by some groups that recommend initiation of breast cancer screening beginning 8 years after radiation or at age 25, whichever occurs last. Drawing primarily from data derived from breast cancer surveillance research in young women with BRCA mutations, mammography and adjunct MRI has been proposed as the optimal surveillance measures for childhood cancer survivors at high risk for breast cancer.

Similarly, there is evidence that that childhood cancer survivors develop colorectal cancer more frequently and at a younger age than the general population, but the median age of onset is not well established. Radiation dose and the volume of bowel in the treatment field are important contributors to risk, and emerging evidence has implicated specific chemotherapeutic agents such as procarbazine and cisplatin. Until there is better understanding of the natural history of gastrointestinal carcinogenesis predisposed by cancer treatment, early initiation of colorectal cancer surveillance has been recommended only for those demonstrated to be at highest risk (eg, those treated with abdominal, pelvic and/or spinal radiation fields at doses of 30 Gy or more). In this high-risk group, initiation of surveillance with colonoscopy at age 35 or at 10 years post radiation, whichever occurs last, is advised by some groups, with the frequency of follow-up guided by the findings.

In general, pediatric oncology late effects researchers are trying to be very judicious about issuing recommendations for screening for adult onset cancers in long-term survivors, as these interventions may be associated with harms and risks to individual patients. My bias is to try not to overscreen. But when health outcomes research demonstrates compelling data to characterize a high-risk group and detection of disease at early stages is associated with better outcomes, initiation of cancer screening may be reasonable. For most adult-onset cancer, limited information is available to appropriately evaluate the benefits — as well as the harms and risks — of screening, which underscores the importance of continued research in this area.

 

Melissa M. Hudson, MD, is director of the Cancer Survivorship Division and co-leader of the Cancer Prevention & Control Program at St. Jude Children’s Research Hospital. She can be reached at Cancer Survivorship, MS 735, Room S-6046,St. Jude Children’s Research Hospital, 262 Danny Thomas Place, Memphis, TN 38105-3678; email: melissa.hudson@stjude.org. Disclosure: Hudson reports no relevant financial disclosures.

 

Until additional evidence is available, surveillance should follow guidelines for non-cancer survivors unless they are a mutation carrier.

Survivors of adult cancers tend to be at a higher risk for the development of future, additional primary cancers than their non-cancer peers. This can be attributed to their underlying intrinsic predisposition to cancer development (eg, genetics or environmental exposure), as well as to cancer treatment. The magnitude and nature of this intrinsic risk is best defined in mutation carriers. However, for non-mutation carriers, evidence is limited in terms of specifics. Furthermore, screening for additional primary cancers might cause distress to cancer survivors due to out-of-pocket costs, the need for time away from work, radiation exposure and fear of another cancer diagnosis.

Health maintenance by screening for non-cancerous conditions (eg, thyroid dysfunction, hyperlipidemia and hypertension) often takes a backseat when cancer patients undergo acute therapies. This should be resumed at some point in the survivorship phase, particularly as prior cancer treatments can negatively affect their risk for these conditions, and these conditions can negatively affect future risk of recurrent or additional primary cancers. Lastly, literature has shown that many cancer survivors suffer from a multitude of clustering, interacting post-treatment concerns (emotional, physical and practical). Cancer survivors are at risk for abnormal healthy behaviors, including those related to diet, exercise, and alcohol and/or nicotine use.

Screening for cancer survivors’ concerns and lifestyles is important, as it can affect their quality of life, ability to work or adhere to self-administered cancer therapies. It also can affect cancer-specific survival and OS.

To conclude, until we have additional evidence directing a more personalized approach, primary care providers and oncologists need to collaborate throughout a cancer patient’s treatment and survivorship phases to allow for consistent, appropriate, and ongoing monitoring of and support for cancer treatment tolerability/recovery, health behaviors and maintenance, and cancer surveillance that follows national age- and gender-appropriate guidelines for non-cancer survivors. Those who are mutation carriers should, in addition, follow the recommendations as made by a cancer genetic counselor.

 

G. van Londen, MD, MS, is director of the University of Pittsburgh Medical Center’s Cancer LiveWell Survivorship Programs, as well as an assistant professor of medicine in the University of Pittsburgh School of Medicine. She can be reached at Hillman Cancer Center, S. 140 Cooper Pavilion, 5115 Centre Ave., Pittsburgh, PA 15232. Disclosure: van Londen reports no relevant financial disclosures.