Study: Declining use of chemotherapy for early-stage breast cancer: examining oncologist recommendations

IN-DEPTH REVIEW OF RESEARCH
Study background:

Chemotherapy benefits cancer treatment in some, but not all cases. Furthermore, it can cause significant side effects—the balance of benefit and risk of this treatment has changed with greater knowledge in this field, and in particular, our expanding ability to predict outcomes from tumor testing.

Increasingly, oncologists are personalizing cancer care, tailoring treatment to an individual's situation. This study is important because it documents the decline of chemotherapy in the face of unchanged guidelines for care, and oncologists’ use of additional information, including recurrence scores and patient preference, to shape their recommendations for chemotherapy.

Oncologists recommend treatment before or after surgery to remove tumors, based on size and stage, and whether cancer cells are present in the lymph nodes. Patients who have cancer cells in the lymph nodes as well as the primary tumor site are generally considered to have more advanced cancer.

Genomic testing of tumors has improved the ability of oncologists to predict the chance of recurrence (a reappearance of cancer after surgery and treatment). A patient’s recurrence score (RS) is a measure of this prediction:

• 0-18 is considered low and is associated with a lower chance of recurrence.
• 19-30 is intermediate and more nuanced.

Another XRAYS review examined results from the TAILORx trial. This study found that women with hormone receptor-positive, HER2-negative, node-negative early-stage breast cancer with Oncotype DX recurrence scores of 11-25 did as well on hormonal therapy alone as those given hormonal therapy plus chemotherapy. Women age 50 or younger with Oncotype DX recurrence scores of 16-25 who received chemotherapy plus hormonal therapy had lower rates of distant recurrence than those who received hormonal therapy alone.

• 31 or greater is considered high and is associated with an increased risk of recurrence.

The National Comprehensive Cancer Network publishes chemotherapy guidelines for estrogen receptor-positive tumors. These are baseline guidelines; oncologists have the latitude to adjust their recommendations to individual patient's circumstances.

Researchers of this study wanted to know:

How chemotherapy use and oncologists' recommendations have changed over time.

Populations looked at in this study:

Women with early-stage breast cancer and their attending oncologists were surveyed for information about chemotherapy use.

The 5,080 women surveyed were identified from two SEER registries in Georgia and Los Angeles between 2013 and 2015. Selection was made based on date of diagnosis; tumor characteristics, including size, stage, grade, estrogen receptor and HER2 status, and whether patients had recurrence score testing.

The women ranged in age from 20-79 years and were diagnosed with stage 0-2 breast cancer. The group of women were 56% white, 16% black, 17% Hispanic and 8% Asian. Among those surveyed, 2,926 met the study criteria with stage 1 or 2 estrogen receptor-positive, Her2-negative breast cancer. (Stage 0 breast cancer patients were excluded, as chemotherapy is not standardly recommended.)

Among 504 oncologists identified for these patients, 304 completed surveys about their responses to node-negative and node-positive case scenarios.

Study findings: 

Chemotherapy use among women with early-stage breast cancer:

Women with stage 1 or 2 estrogen receptor-positive, Her2-negative breast cancer were surveyed about their age at diagnosis, whether they met with an oncologist, received chemotherapy and whether their oncologist recommended chemotherapy.

Among these women, chemotherapy use declined from 34% in 2013 to 21% in 2015. During this time there were no major changes in U.S. guidelines.

• For women with node-negative cancer, chemotherapy used declined from 27% to 14%.
• For women with node-positive cancer, chemotherapy used declined from 81% to 64%.

Estimating recurrence is conducted with a 21-gene panel test of tumors. Low tumor recurrence scores (RS) indicate that there is a low probability of recurrence. High tumor RS indicates that there is a higher probability of recurrence. Use of RS changed the frequency that chemotherapy was used for node-positive but not node-negative patients:

• Among node-positive patients, use of recurrence scores accounted for one-third of the decline in chemotherapy; even if patients were node-positive or if their tumor recurrence scores were low, there was a trend to omit chemotherapy.
• Among node-negative patients, there was no change in use of recurrence scores.

Recommendations of oncologist in response to case scenarios:

Researchers presented oncologists with two hypothetical case scenarios:

1) A 60-year-old, postmenopausal patient with a 0.7 mm tumor who was node-negative.

2) A 48-year-old, premenopausal patient with a 2.2 cm tumor who was node positive.

Oncologists were asked for their recommendations regarding chemotherapy:

• based on the hypothetical case history alone.
• based on case history and added information about patient preferences that disagreed with the oncologist's initial recommendation.
• based on added information about recurrence score results.

For the first case (a more favorable disease situation):

• 9.3 % of oncologists recommended chemotherapy.
  • Two-thirds (74%) of oncologists would order genomic testing of the tumor before making a decision.
• 13% of oncologists recommended chemotherapy when patients indicated they preferred it.
  • More oncologists (87%) would order genomic testing of the tumor before making a decision.
• 96% of oncologists recommended chemotherapy, substantially more than for the same patient without RS testing, when the patient's tumor had high RS of 34 (an unfavorable prognosis).

For the second case (a less favorable disease situation):

• 99.6% of oncologists recommended chemotherapy.
  • Only 17% would order genomic testing of the tumor before making a decision.
• Slightly fewer (98.9%) oncologists recommended chemotherapy when patients indicated they preferred to no chemotherapy.
  • 67% of oncologists would order genomic testing of the tumor before making a decision.
• 56% of oncologists recommended chemotherapy when the patient's tumor had low RS of 16 (favorable prognosis), substantially fewer than for the same patient without RS testing.

Overall, oncologists considered patient preferences when ordering tests or recommending treatment, particularly if patient wishes disagreed with their original recommendations. Oncologists used recurrence scoring to help reconcile differences in patients’ wishes and their recommendations.  Whether this trend is warranted remains to be seen.

Limitations:

This study has a few limitations. First, the data on chemotherapy use relies on patient reports rather than more direct measures (e.g. patient medical records).

The number of node-positive cases were smaller (420 or 14% of all cases) than node-negative cases (over 2,000). There was no detail on lymph node number, size, or specific chemotherapy agents recommended. More details would allow a more nuanced understanding of oncologist recommendations.

The study took place over a relatively short period (two years) in only 2 geographic locations, and oncologists were surveyed only once. It is unclear if oncologists' recommendations would differ depending on geographic location or whether recommendations would change more (or less) with increased time interval and more reflection.

Conclusions:

Chemotherapy use substantially and significantly decreased over the 2-year evaluation period, despite a lack of changes in national guidelines. The decline in use may reflect improvements in breast cancer outcomes for early-stage disease, increasingly personalized care and concerns about overtreatment. Oncologists were responsive to patient preferences in their recommendations about chemotherapy, and in particular, their willingness to pursue genomic testing of tumors to clarify prognosis. In the face of genomic test results that ran counter to initial chemotherapy plans, many oncologists adjusted their recommendations.

Share your thoughts on this XRAYS article by taking our brief survey.

Posted 8/21/18