Complementary and alternative medicine (CAM) supplements in cancer outpatients: analyses of usage and of interaction risks with cancer treatment

Complementary and alternative medicine (CAM) comprises all modalities that are used instead of (alternative medicine) or in addition to (complementary medicine) conventional and more stringently evidence-based medicine by patients (NCCIH 2018; Teichfischer and Muenstedt 2011). These include, for example, the use of herbal substances, non-prescribed use of supplements such as vitamins or minerals, homeopathy, Chinese medicine, massage, acupuncture, prayer, and more.

The use of CAM is widespread among cancer patients. Exact frequencies vary between different studies, e.g., 29% Firkins et al. (2018), 34% Wode et al. (2019), 36% Molassiotis et al. (2005), 40% Horneber et al. (2012), 49% Berretta et al. (2017), 59% Micke et al. (2009). Younger age, female gender, and higher educational level (Micke et al. 2009; Molassiotis et al. 2005; Wode et al. 2019), as well as breast cancer diagnosis (Micke et al. 2009), are discussed as predictive parameters for CAM use. Since the 1970s, the number of patients using CAM has been increasing (Horneber et al. 2012).

Consuming supplements including vitamins, minerals, and plant extracts is the most commonly used CAM modality in Europe and the United States (Alsanad et al. 2016; McCune et al. 2004; Micke et al. 2009; Molassiotis et al. 2005; Naing et al. 2011), which can potentially interact with conventional cancer therapy. A high prevalence of potential CAM–drug interactions with conventional anticancer drugs has been reported in several studies, but numbers are heterogeneous: e.g., 55% Firkins et al. (2018), 65% Zeller et al. (2013), 85% Loquai et al. (2017).

The aim of our study was to analyze the use of CAM supplements by calculating associations with demographic data and to identify possible predictors, as well as to analyze and compile potential interactions of CAM substances with conventional cancer therapy in terms of their probabilities.

Critical issues with CAM and its potential interactions are the lack of in vivo data and clinical studies investigating the clinical relevance of CAM substance interactions. The potential for substance interactions often can only be estimated from in vitro experiments or murine models (Huebner 2012).

In fact, most if not all patients want to avoid harming themselves when using complementary medicine. In a study of outpatient cancer patients by McCune et al. (2004), more than 85% conceded that they would stop using a CAM drug or ask their physician for advice if interactions were known. However, potential interactions are still common, presumably reflecting patients’ and physicians’ unawareness regarding the risks of natural substances, most of which are over-the-counter drugs distributed not only by pharmacies but also via the internet.

One hundred fifteen patients participated in our study. Demographic data are presented in Table 1. The mean age was 61 years (SD = 13.3). 40.9% of patients were male. The most common diagnosis categories (Table 2) were breast cancer (n = 25), other gynecological cancers (n = 15) such as ovarian cancer, cervical cancer, or endometrial cancer, multiple myeloma (n = 15), and leukemia (n = 10).

Non-prescribed supplements, which were considered CAM supplements, were taken by 43 patients (37.4%). These patients took a total of 117 CAM compounds. Twenty-two patients consumed vitamin supplements, 24 supplemented minerals including trace elements, 19 used certain food or plant extracts like Brazil nuts, Chinese herbs, ginger, medicinal mushrooms, mistletoe, teas, turmeric, spirulina, or others, and 9 patients consumed other processed CAM compounds like homeopathy or probiotics. Seven of the 43 CAM supplements using patients indicated that physicians were the only ones who recommended the use of CAM to them.

Table 2 shows CAM use by cancer diagnosis type. The rate of patients using CAM varied between the individual diagnosis categories. While none of the ten leukemia patients took additional substances, a high rate of patients using CAM substances was found in myeloma patients. Eight out of 15 patients with multiple myeloma (53.3%) indicated using CAM supplements. Gynecologic cancer patients used CAM substances in 15 out of 40 cases (37.5%) what was about average. More women stated using CAM supplements (29 out of 68, 42.6%) than men (14 out of 47, 29.8%). The 29 women using CAM supplements took 83 different supplements (2.9 in average per patient, SD = 2.6), which included many different active ingredients whereas the 14 men using CAM supplements took 34 different CAM supplements in total (2.5 in average per patient, SD = 1.6).

The use of CAM supplements was statistically analyzed as shown in Table 3. Correlations were calculated for use (yes/no) and number of CAM supplements used with age (age over 61 years), gender, marital status, having children, having minor children, school leaving qualification, type of cancer diagnosis, time since initial cancer diagnosis (time longer than 1 year), and with the number of drugs prescribed (n > 9). Significant (p < 0.050) associations were found regarding the use of CAM supplements: Individuals older than 61 years were less likely to use CAM than younger patients (p = 0.020, φc = 0.229). When the time since diagnosis was longer than 1 year, the use of CAM supplements was significantly more likely (p = 0.006, φc = 0.264). A model of binary logistic regression including these two parameters to explain CAM use also showed significant results (Table 3), while the overall model fit was low (χ² = 12.9). However, the number of different CAM compounds consumed did not correlate with any of these parameters.

Potential CAM–drug interactions between CAM ingredients and conventional cancer treatment revealed in 22 of the 43 CAM supplements using patients (51.2%). There was a higher risk of CAM–drug interactions with anticancer drugs in patients taking a higher number of CAM compounds (p < 0.001, OR = 3.660, CI95% [2.183, 6.137]), but not when they received more drugs for cancer treatment (p = 0.648, OR = 0.874, CI95% [0.491, 1.556]). The overall model fit was χ² = 53.0.

The potential risks of interactions with conventional cancer therapy were assessed based on literature data. For all active CAM ingredients taken by patients in our study, the potential interactions with conventional anticancer drugs are shown in Table 4. For comprehensiveness, regimes that were not prescribed to the patients of our study are also included. The actual occurrence in patients was not investigated in our study.

37% of all patients reported using CAM supplements. This is in line with Loquai et al. (2016) who found that 34% used CAM supplements (biological-based CAM), and Alsanad et al. (2016) who reported a rate of 34% of all patients using herbal or dietary supplements. Molassiotis et al. (2005) found that 36% of all included European cancer patients used CAM of any modality with a range of 15% to 73% across countries, while CAM supplements such as herbal medicine or dietary supplements are the most frequently used CAM modalities in Europe and the United States (Alsanad et al. 2016; McCune et al. 2004; Micke et al. 2009; Molassiotis et al. 2005; Naing et al. 2011; Zeller et al. 2013). Data on the use of not only CAM supplements but all CAM modalities vary between regions, as reported even between European countries (Molassiotis et al. 2005). For Saudi Arabia, a percentage of 69.9% CAM use is reported among cancer patients mainly using religious CAM modalities and camel products (Abuelgasim et al. 2018).

Different demographic parameters of the patients are discussed in the literature as predictors of CAM use. The most commonly considered ones are female gender, younger age, higher education level, and breast cancer as the type of cancer diagnosis (Micke et al. 2009; Molassiotis et al. 2005; Naing et al. 2011; Richardson et al. 2000; Wode et al. 2019). Table 5 shows a selection of studies, which showed different results regarding the significance of various parameters within this context.

Regarding younger age, our study also showed a significant association with CAM use. In addition, our results suggest that a longer period of time since initial diagnosis is also suitable as a predictor. Among the three most frequently mentioned reasons for using CAM in cancer patients are the attempt to reduce side effects of conventional therapy and the desire to become more active against the cancer disease (Huebner et al. 2014). This may explain why CAM use increases with the duration of disease: when side effects of conventional therapy first appear in the course of the disease, or when patients’ desire to influence the course of their cancer themselves increases as the disease progresses. Although the number of patients using CAM supplements in our study differed within the diagnosis categories, no significant association was found between type of cancer diagnosis and CAM use. A higher rate (64–76%) of CAM users among breast cancer patients described by some authors (Huebner et al. 2014; Zeller et al. 2013) could not be confirmed. In fact, breast cancer patients used CAM slightly less frequently than the average of all patients (36% vs. 37%). Random influences might affect our results, as the total number of breast cancer patients was small (n = 25) compared to other authors (Micke et al. 2009; Molassiotis et al. 2005). Nevertheless, even with these authors, the data are inconclusive. The international multicenter study by Molassiotis et al. identified cancer entities others than breast cancer with higher rates of CAM using patients (2005). In contrast to other authors, we could not determine gender as a reliable predictor of CAM use (Micke et al. 2009; Molassiotis et al. 2005; Naing et al. 2011; Richardson et al. 2000; Wode et al. 2019). As described by Richardson et al. (2000) too, we could not detect a significant association with marital status. Having one or more children or minor children as well as the consumption of a high number of conventional drugs did not prove to be suitable predictors either.

Yet, the use of individual parameters as predictors of CAM usage should be treated with caution. Calculated effect sizes, when indicated, were small and the models could only explain a small part between 6 and 26% of the difference between the groups of CAM users and non-CAM users (Molassiotis et al. 2005; Richardson et al. 2000; Wode et al. 2019), as was the case also in our study. In all patients, regardless of age, gender, cancer diagnosis, etc., CAM use should be investigated and documented by physicians in charge to identify potential interactions with conventional cancer therapy.

More than half (51%) of all patients using CAM supplements were at risk of CAM–drug interactions involving their cancer treatment prescribed by physicians. This is less than has been reported by other authors who used a similar classification system on potential interactions as we did. Up to 65% (Zeller et al. 2013) and even 85% (Loquai et al. 2017) are reported elsewhere. A lower number of interactions compared with the studies mentioned above might occur if the patients in our study took fewer CAM supplements per person than in the other study. Other authors focused on a particular source of CAM–drug interactions, such as interactions via CYP enzymes and Pgp (P-glycoprotein) interactions (Engdal et al. 2009) and examined lower interaction rates, as might be expected. In addition, authors of different studies evaluate the likelihood of CAM–drug interactions differently, so results based on different classifications (e.g., Engdal et al. 2009; Lee et al. 2006; McCune et al. 2004; Werneke et al. 2004) appear difficult to compare.

Our analyses showed that the potential of CAM–drug interactions with cancer therapy was significantly related only to the number of CAM supplements taken but, to our surprise, not to the number of anticancer medication prescribed by physicians. This could be explained partly by the fact that CAM substances have a rather dichotomous pattern of potential interactions, according to our research: in many cases, CAM substances either had a very wide range of potential interactions, such as curcuma longa or vitamin C, or interactions were generally unlikely, as is the case with most minerals. Therefore, the number of CAM products taken may be critical because as the number of CAM substances increases, the likelihood increases that there will be a CAM supplement among them that has a broad spectrum of interactions. A smaller number of CAM products taken is more likely to be products for which no potential interactions have been explored. Conventional drugs, on the other hand, seem to have a more evenly distributed interaction spectrum, so how many are taken is less relevant to the overall potential for interactions (yes/no).

For some CAM substances, bioavailability is high and substantial serum concentrations of the active ingredients can be reached easily, which increases the risk of interactions. Taking vitamin C as an example, several studies have shown reduced effects of chemotherapy regarding a variety of therapeutic regimes: anthracyclines, bortezomib, bleomycin, cisplatin, vincristine, methotrexate, and imatinib (Heaney et al. 2008; Perrone et al. 2009; Pohl and Reidy 1989; Zeller et al. 2013). In addition, reduced toxicity of various chemotherapeutic drugs with the use of antioxidants is discussed, such as for oligomeric proanthocyanidins (OPC) with regard to anthracyclines or cyclophosphamide (Joshi et al. 2000; Li et al. 2010; Sharma et al. 2004), while the influence on the efficacy of cytostatic therapy has not yet been sufficiently studied (Huebner 2012 p342). A recent large cohort study by Jung et al. (2019) showed that taking antioxidant supplements during radio- or chemotherapy is associated with higher mortality as well as a reduced recurrence-free survival.

Interactions are difficult to denote in individual patients, as their influence on the course of the disease cannot be proven in most cases. To protect patients from such interactions, proactive recommendations not to use supplements which might entail interactions seem appropriate. Therefore, physicians should strive to know about supplement use of the patients they care for. Only about one-third (33–36%) of all patients consuming CAM supplements talked to their physician about it (Firkins et al. 2018; Kennedy 2005). Consideration and documentation of CAM supplements by treating physicians seems important and necessary to identify possible interactions. This requires knowledge of the benefits, risks, and potential interactions of CAM supplements by physicians and pharmacists. A study published in 2009 investigating drug interactions with CAM supplements complained of a lack of literature on potential interactions in 48% of the herbal remedies used by patients (Engdal et al. 2009). Although many more studies have been published since then, most of the data still bases on mouse models or in vitro experiments (e.g., Huebner 2012). Clinical trials, data from large cohorts and registries, and more efforts to obtain reliable information are essential to effectively counsel patients based on human data.

In a U.S. study, more than 50% of CAM using patients indicated that CAM supplements were important for their well-being and health (Kennedy 2005). On the other hand, the patients do not want to harm themselves. Education about the possible consequences of interactions is important. Moreover, patients’ desire to be active, participate in treatment, and reduce side effects while improving quality of life should be acknowledged, and physicians should offer advice on safe CAM methods, which may be certain supplements but also healthy lifestyle, nutrition, and physical activity.

Interactions of conventional cancer therapy with over-the-counter CAM supplements are often underestimated and yet insufficiently researched. High doses of supplements in the form of extracts and concentrates should be avoided, especially in cancer patients, if there is no proven medical indication for their use, e.g., a deficiency of a micronutrient. Overdoses should be avoided in all cases. Physicians should address risks and document the use of CAM supplements in patient records and check for interactions. Further studies are needed for a variety of CAM supplements regarding benefits and risks, such as interaction risks.