Thank you for allowing me to present the results of SWOG 1200. We’ve heard earlier today and throughout this conference about the efficacy of aromatase inhibitors for women with hormone sensitive breast cancer, it works both in the adjuvant setting in women with metastatic disease  and for the prevention of breast cancer, but we also know that it doesn’t work if you don’t take it. Non-compliance is a major problem amongst women taking hormonal therapy. While non-compliance is multi-factorial, one of the main reasons why women discontinue their treatment early Is due to arthralgias, or joint discomfort. We were interested in a non-pharmacologic modality for assessing whether or not we could control these symptoms.

We looked at acupuncture, which is a popular modality for the treatment of a variety of medical conditions and several single institution studies have suggested that acupuncture may be beneficial for AI arthralgias. However, differences in methodology, small sample size and duration have raised questions about the benefit.

SWOG 1200 was designed to evaluate women with AI induced joint pain. It randomised women two to one to one, to true acupuncture given twice a week for six weeks, sham acupuncture given twice a week for six weeks, or waitlist control. After six weeks was the primary endpoint. We then went on to test whether or not a maintenance intervention would maintain the effect. Women received once a week acupuncture for six weeks, once a week sham for six weeks, or maintained on the waitlist control. We also wanted to know the duration of the effect so we followed women for an additional twelve weeks with no treatment in all three groups.

We had two primary endpoints testing true versus sham versus true versus waitlist control. The reason is that sham can sometimes have a physiologic effect. Eligibility was stage one to three hormone sensitive breast cancer, they needed to be on an aromatase inhibitor for at least thirty days. As I said before they needed a score of at least three on the worst pain item of the Brief Pain Inventory. Their symptoms had to have started or increased since starting an aromatase inhibitor. They could not take opioids or corticosteroids; no alternative therapies or physical therapy for joint pain for the 28 days prior. Patients could not have prior acupuncture treatment for their joint pain at any time but they were allowed to have had acupuncture for other reasons as long as it was greater than twelve months prior.

The intervention was administered by licensed acupuncturists, it consisted of a standard traditional point prescription to reduce pain and decrease stress, as well as a protocol that was tailored to the most painful joints. The sham included shallow needle insertion using thinner and shorter needles at non-acupoints. For the joint specific sham, protocol was within proximity to the specified area.

The baseline characteristics between the three groups were identical. Essentially, the median age was 60, the majority of patients were white, half had prior chemotherapy. The median time on an AI was just over a year, and about 20% had prior acupuncture for other conditions. The median baseline score was 6.5.

Here are the primary results for the worst pain score on the Brief Pain Inventory. There was a statistical difference between the scores for true acupuncture versus sham acupuncture and true acupuncture versus waitlist control, and no difference at six weeks between sham and waitlist control. We also looked at the percentage of women with a two point change, which we considered clinically meaningful. Other studies have looked at a 50% reduction. You can see 58% of women in the true acupuncture arm had a two point change compared to 31% in the sham group and 30% in the waitlist control. These results were highly statistically significant, and we found similar results looking at a 50% reduction.

Here are the results over time with a linear mixed model, assessing all of our assessments over the 24-week period. You can see that over time true acupuncture was better than sham, and true acupuncture was better than waitlist control. Looking at other secondary endpoints with either the Brief Pain Inventory average pain, worst stiffness, WOMAC score, looking at pain in knees and hips specifically, or scale looking at hand pain, all of the results were identical, true was better than sham, and true was better than waitlist control. Looking at the linear mixed model for other endpoints you can see essentially the same trend over time with the results being maintained at 24 weeks. With regard to side effects, there was only one grade 2 in both the sham and true arm, which was presyncope, the difference between the arms was with regard to grade 1 bruising, 47% versus 25% in the true versus the sham arms.

In conclusion, we have shown consistently, with multiple measures assessing pain and stiffness, that true acupuncture generated better outcomes than either control group. Transitioning from twice-a-week to once-a-week acupuncture maintained the effect, and the intervention effects persisted twelve weeks following completion of the intervention. The toxicity was minimal and limited to grade 1 bruising. The clinical implications are that acupuncture provides a non-pharmacologic option that can improve symptoms and possibly increase aromatase inhibitor adherence and possibly subsequent breast cancer outcomes if patients can stay on their medication.

For patients reluctant to take prescription medications to control pain that can result in other side effects, acupuncture provides a safe and effective alternative. Identification of non-opioid options such as acupuncture for pain control is a public health priority. The cost of the intervention, twelve weeks or eighteen sessions, was about $1,200 - $1,500. We feel there is now sufficient evidence to support insurance coverage for acupuncture for the treatment of aromatase inhibitor arthralgias. Thank you.

Thank you Dawn. Are there any variables that could affect these results and if so were they controlled for? I’m speaking out of ignorance here, things like BMI, exercise?

Those are excellent questions, BMI was equal in the three different groups. We didn’t measure exercise activity during the intervention, other studies have shown that exercise can be beneficial.