Homeopathy for breast cancer surgery? Isn’t it bad enough that the patient has cancer and needs a mastectomy? [Respectful Insolence]

I like to refer to homeopathy as The One Quackery To Rule Them All, so much so that I almost always call it that within the first two paragraphs of any post I write about some tasty bit of homeopathy pseudoscience. It’s also a wonderful tool for teaching critical thinking because it’s easy to explain and people grasp intuitively why homeopathy is pseudoscience when it’s explained properly to them. Basically, it’s because of homeopathy’s two laws. The first is the Law Similars, which states that, relieve a symptom, you must use something that causes the symptom. It’s nonsense. There’s no science behind a general rule like that. Then, there’s the second law, the law of infinitesimals, which states that a remedy gets stronger with dilution. That’s why homeopaths serially dilute their remedies—with intense shaking between each step of dilution (homeopaths claim that the shaking, or succussion, is absolutely essential)—to the point where it is unlikely that there is a single molecule of the remedy left. For instance, a 30C homeopathic remedy is a remedy that’s been diluted 100-fold thirty times, which results in a total dilution of 10^-60. Given that Avogadro’s number (the number of molecules in a mole of a chemical) is roughly 6 x 10²³, that means a 30C dilution is over 10³⁶ more than Avogadro’s number. Most people think that homeopathy is nothing more than herbal medicine and have a hard time believing it when I tell them what homeopathy really is.

All of this why skeptics frequently say that homeopathy is water, using that observation as the basis for jokes. Now, it’s true that some “weaker” (i.e., less dilute) homeopathic remedies haven’t been diluted enough times to dilute the original remedy to nothing (i.e., there’s still something left), but most homeopathic remedies are, in fact, water or ethanol or whatever homeopaths used to dilute the remedy, usually pressed into sugar pills. A purer placebo, it is hard to imagine.

That’s why it never ceases to amaze me when actual surgeons—even worse, breast cancer surgeons, my peeps!—forget everything they were taught in undergraduate and medical school about chemistry and physics and think that it’s worthwhile to test a homeopathic remedy in a randomized clinical trial for breast surgery. After all, in such a trial, no matter what you choose as the placebo control, you’re testing placebo versus placebo. In this case, it’s surgeons in Italy, along with what appear to be basic scientists, testing the effect of a homeopathic remedy on healing in breast surgery. The remedy being tested is even more clearly water than even a 30C remedy. This time, the remedy being tested is Arnica montana, which is also known as wolf’s bane, a flowering plant that is a member of the sunflower family. It’s a common homeopathic remedy, and in this case it’s really, really, really dilute:

A. montana 1000 K was in the form of drops in 30% hydro alcoholic solution, with sublingual absorption. The 1000 K dilution is a very high homeopathic dilution produced according to the French Pharmacopoeia, starting from a mother tincture (complete plant extract) which undergoes 1000 steps of successive dilution and shaking in a 30% water–alcohol solution. The mother tincture of the product used in this study contained a minimum content of 0.04% of sesquiterpenes (expressed as dihydrohelenanin tiglate). According to the Korsakovian dilution method, the same flask was used for the entire preparation: At each step, the flask was emptied of most of the liquid but not dried, and then immediately filled again with the solvent. Because the residual volume after emptying the flask certainly cannot be >10% of the total used in each dilution, the 1000th K dilution is sufficiently high that it certainly does not contain toxic amounts of the plant. The placebo drops were identical in appearance to the active drops, but included only 30% hydro alcoholic solution without any homeopathic dilution. The studied homeopathic treatment and the placebo were manufactured and supplied in strictly identical packaging by Laboratoires Boiron S.r.l. (Segrate, Milan, Italy).

So what we’re talking about is one thousand 100-fold serial dilutions, resulting in a final 10^-2000 dilution. This is madness, as it’s estimated that there are only around 10⁸⁰ particles in the known universe. Seriously, all that is left there is 30% alcohol. So basically the study I’m examining is testing the effect of 30% alcohol on wound healing in breast surgery. You’ll see what I mean. Here’s the rationale:

Surgery is not free from significant complications, despite advancement in surgical techniques and in perioperative care. One of the major complications, particularly for breast and soft tissues surgery, is post-operative bleeding which represents a critical and, in some cases, lethal risk factor [1,2]. Moreover, post-operative bleeding after breast surgery causes a severe discomfort to the patient, entailing the need for surgical re-intervention and sometimes blood transfusions. Another frequent complication of breast surgery is seroma, which often requires numerous accesses for outpatient drainage, finally resulting in a delay of adjuvant therapy administration and a significant psychological burden [3].

All of which is true, but doesn’t justify this:

The risk of blood loss or seroma could be limited by identifying and then correcting potential triggers, but although this topic has been extensively studied, both hemorrhage and seroma are still largely reported [4,5]. Homeopathy could provide remedies based on compounds with anti-hemorrhagic and anti-inflammatory properties, but their use in diluted homeopathic solutions has been poorly investigated [6-9]. Indeed, some homeopathic remedies have been already evaluated in various surgical settings, however showing controversial effects [10-17]. Arnica montana is a plant belonging to the Compositae family and it grows in East and Central Europe [18]. In particular, its anti-inflammatory action is linked to the lactone helenalin that seems to be involved in the inhibition of the pro-inflammatory transcription factor nuclear factor kappa B (NF-κβ) [19-27]. A. montana has been recently evaluated in several surgical settings such as esthetic and orthopedic surgeries, but currently, no single study is available in literature on the effects of this homeopathic remedy in breast cancer surgery. In the present study, a homeopathic preparation of A. montana 1000 Korsakovian dilution (1000 K) was administered preoperatively and postoperatively in a placebo-controlled, double-blind clinical trial to patients undergoing unilateral total mastectomy, with or without reconstruction, to evaluate any favorable or adverse effect on post-operative blood and seroma collection from surgical drains.

OK, let me back up and do a bit of explaining here. After a mastectomy, we generally place a drain under the skin flap in order to collect the drainage, which starts out bloody but clears up until it’s serum, usually pink with a little bit of blood mixed in. When the drainage decreases to below a certain volume per day, the drains are removed. Patients don’t like drains because they’re inconvenient, messy, and can clog, but without drains, it’s almost inevitable that there will accumulate something that’s called a seroma, which is a collection of that serous fluid under the skin flap left after the mastectomy. Sometimes two drains are placed, usually when the patient has her axillary lymph nodes (the lymph nodes under the arm) removed, because the cut lymph vessels also leak lymph and can result in a seroma.

In this study, the investigators studied the effect of 1000K A. montana compared to placebo on the following endpoints:

• Duration of drainage, which was calculated from day 1 to the day of drainage removal
• Time to reach a collected volume below 10 ml
• Self-evaluation of pain, measured by VAS after surgery
• The average time of hospitalization after surgery, which was measured in days following surgery
• The presence of bruising and hematomas or breast swelling, measured both in terms of presence versus absence and of description or size
• The differences between the volume collected on day 1 and the volume collected in each of the following days
• Possible adverse reactions to treatment.

The patients were randomized to one of two grsoups:

The patients, according to the randomization list, received sublingually a dose of five drops of A. montana 1000 K 3 times a day, or placebo with identical times of administration, from 1st day before the surgery until the 4th day after the surgery (in total 6 days).

The results were analyzed using three models:

• Model 1: Includes the treatment;
• Model 2: Includes the treatment and the volume collected on the day of the intervention;
• Model 3: Includes the treatment, the volume collected on the day of surgery, and patient weight.

Overall, 53 women were enrolled and randomized (26 in the A. montana group and 27 in the placebo group). One defect in this paper is that it didn’t show a proper CONSORT flow diagram, which is generally mandatory for any halfway decent clinical trial. Be that as it may, what did the investigators find?

Nothing that spectacular:

The per-protocol analysis revealed a lower mean volume of blood and serum collected in drainages with A. montana (−94.40 ml; 95% confidence interval [CI]: 22.48-211.28; P = 0.11). A regression model including treatment, volume collected in the drainage on the day of surgery, and patient weight showed a statistically significant difference in favor of A. montana (−106.28 ml; 95% CI: 9.45-203.11; P = 0.03). Volumes collected on the day of surgery and the following days were significantly lower with A. montana at days 2 (P = 0.033) and 3 (P = 0.0223). Secondary end points have not revealed significant differences.

The first thing to notice is that there really wasn’t a lower mean volume in blood and serum collected in the patients treated with the homeopathic remedy. the p-value was only 0.11. Now, apparently the criterion used by these surgeons for removing the drains was 10 ml/24 hours, because that was the endpoint the investigators looked at. My criterion is a bit more liberal. Be that as it may, the results are not impressive. For instance, take a look at this graph, which shows the change in drainage after day 1:

Yes, there does appear to be a slightly higher amount of drainage in the placebo group. It doesn’t matter though. When it comes to breast surgery, what matters is not so much how much drainage is collected through the drains, but how many days before the drains can be removed. That number is the same for the placebo and homepathic A. montana groups.

To drive home the point, look at this graph, which shows the time for each group to reach a <10 ml of drainage>

Basically, there was no difference, as I would expect. Certainly there is no clinically significant difference. Even the authors seem to know that:

In this study, which to our knowledge is the first to test the effect of A. montana in breast cancer surgery, we have measured the efficacy of the homeopathic treatment versus placebo in reducing the post-operative bleeding and seroma production in patients subjected to total mastectomy. The difference in the volume of blood and serum collected by drainages between the group of women treated with A. montana and those treated with placebo was statistically significant in the PP dataset, but only in the model which included, in addition to the type of treatment, the volume collected on the day of surgery, and the weight of the patient. Thus, we determine that these two additional variables have a considerable effect on the efficacy of the treatment, and further studies are needed to investigate their specific contributions.

Um, no. Basically you had to keep trying different models until you found one that produced a (barely) statistically significant result. I do have to give the authors props for this unintentionally hilarious observation:

The main limitations of the study were the relatively small sample, the analysis of post-operative recovery, which was limited only to 5 days, and the use of a high Korsakovian dilution which could have negatively impacted on the statistical significance of our findings. Observational studies with different Korsakovian dilutions could be useful to highlight any significant effect of A. montana and further validate these findings.

No. They. Could. Not. Be.

And this statement sums up every homeopathy study ever:

The scientific community often claims that homeopathy effects are not supported by rigorous clinical trials. This study, although reporting statistical significance only for some specific data settings, suggests a reduction in post-operative blood loss and seroma production in a group of women who underwent breast cancer surgery and treated with A. montana 1000 K.

Seriously, this is about as negative a study as can be imagined, just like every other study of homeopathy that’s anywhere near competently designed and carried out. That’s because homeopathy is water, or, in this case, 30% ethanol in water. At the dosage given, it there’s not even enough alcohol there to give you a little buzz. It’s like taking a teaspoon of cough syrup, many of which also have alcohol in them.

Once more, here we have a pseudoscientific and unethical clinical trial that shows…nothing. That’s appropriate, though, because homeopathy is basically nothing.