Effects and Safety of Massage Therapy in Cancer Care
This section reviews research evidence on the safety, feasibility, and effectiveness of MT across the cancer care spectrum in both children and adults and examines the effectiveness and limitations of MT to reduce both cancer- and treatment-related symptoms, such as pain, depression, anxiety, nausea, and fatigue. It also presents the limited evidence on improved biological and immune-response outcomes. It considers the adaptation of treatment sometimes necessary for this population and notes special precautions, such as avoiding radiation sites. In addition, it debunks the myth that MT must be avoided by persons with cancer.
Serious adverse events are rare and massage is generally safe when given by credentialed and experienced practitioners (Corbin 2009; Deng et al. 2009). More than 60 trials provide evidence for the feasibility and safety of massage for children and adults at every phase of the cancer experience. In a systematic review of 10 studies involving 386 patients with cancer, one case report of a skin rash was the only adverse event reported (Wilkinson, Barnes, and Storey 2008). Other anecdotal reports in patients with cancer include headache, lightheadedness, muscle tenderness, or general feeling of unwellness for a day or two following deep tissue massage. Adverse effects can often be prevented with awareness, modification of touch and pressure, and changes to positioning. The Society for Integrative Oncology recommends the use of massage for cancer pain and anxiety, but cautions against the use of deep or intense pressure near cancer lesions or enlarged lymph nodes, radiation field sites, intravenous catheters and medical devices, and anatomic distortions, or in patients with a bleeding tendency (Deng et al. 2009).
It is a myth that massage spreads cancer (Corbin 2005; Ernst 2003). Metastasis of cancer cells is a complex interaction of structural, biochemical, hormonal, immunological, and genetic factors that control cell growth, adhesion, angiogenesis, cell signaling, and mobility. However, inflammation, pain, or sensitivity to pressure are other reasons practitioners should use a gentle touch and avoid direct or deep massage over a surgical or tumor site, or in a limb distal to lymph node removal. Other problems requiring modification of the depth or location of touch include infection, skin irritation, platelet or clotting disorders, bone metastasis, and nausea. Deep massage should be avoided in fields of radiation, since the skin may be fragile and the underlying tissue may be fibrotic or edematous. No oils or lotions should be used on the field of treatment during the course of radiation. Rocking motions should be avoided with patients who are experiencing nausea (Gecsedi 2002). Practicing diligent hand washing and using clean equipment and linens will reduce the risk of infection.
Caution should be used for patients with cancer-related pain. Massage will not resolve pain resulting from pressure of the tumor on surrounding sites, nerve impingement (radiating pain), or bone pain from metastases. Light-touch massage can reduce anxiety or distress, whereas deep massage can worsen the pain by increasing inflammation or causing fragile bones to break.
It is important to assess and adapt touch for each client. In one study, children declined therapeutic massage if they were feeling nauseated, in pain, or ill, but responded positively to light and comforting caress of their arms or legs from their parents (Post-White, Fitzgerald, Savik, et al. 2009). Long, slow, light touch triggers a parasympathetic (relaxing) response, whereas deep massage stimulates the sympathetic nervous system, which increases heart rate and blood pressure and may lead to distress. To provide safe and effective massage, MacDonald (2007) recommends reducing the pressure, slowing the strokes, shortening the session length, and working gently.
Cancer treatment can leave patients feeling lonely and vulnerable. Children may be particularly sensitive to experiencing changes in body image and being unclothed. Asking permission and proceeding gently when touching an amputated limb stump, a bald head, or a scar, even if it is well healed, conveys compassion and respect. Massage to the abdomen can trigger emotional responses. Calm, confident, and loving touch can be restorative and healing as the patient adjusts to a new body image and sense of self. (See chapter 14.)
Efficacy and Effectiveness
Efficacy is the assessment of whether treatment works under controlled conditions in a clinical trial. Controls within the study design reduce bias and eliminate alternate explanations for observed effects, resulting in a cause-and-effect prediction of a given probability. A treatment is efficacious when it proves to be superior to placebo or control conditions. By contrast, effectiveness is the assessment of whether treatment works in a typical clinical setting and is clinically relevant. Both are important to assessing the value of massage to the care of persons with cancer.
Several excellent summaries and systematic reviews evaluate the efficacy of massage research for children and adults with cancer. They include detailed tables of study samples, design, massage techniques, and findings (Ernst 2009; Fellowes, Barnes, and Wilkinson 2008; Hughes et al. 2008; Jane et al. 2008; Myers, Walton, and Small 2008; Russell et al. 2008; Wilkinson, Barnes, and Storey 2008; Beider and Moyer 2007).This chapter summarizes the findings and includes additional research published subsequent to the reviews.
Studies in Adults With Cancer
The most consistent and strongest effect of massage, aromatherapy massage, and foot reflexology in adults with cancer is reduced anxiety (table 17.1). Although most studies measure short-term effects, Wilkinson and colleagues (2007) found patients experienced less anxiety 2 weeks after 4 weekly aromatherapy massage sessions. Future research should determine the length of effects experienced following massage and the clinical relevance is for short-term reduction of anxiety.
Massage reduced depressive symptoms or depressed mood in some patients (table 17.1). Treatment with a single massage therapist over time was more effective than having different therapists in reducing depression in breast cancer survivors (Listing et al. 2009).
In most studies, massage relieved pain immediately after the session (table 17.1). However, some studies found no effect or only selective effects (Weinrich and Weinrich 1990). In longitudinal studies, Listing and colleagues (2009) found pain was reduced after 5 weeks of twice-weekly massage, while Kutner and others (2008) found similar pain-reduction effects for both massage and simple touch. In two other studies, patients used fewer analgesics after receiving massage (Post-White et al. 2003; Wilkie et al. 2000). Massage appears to be most effective for short-term relief of pain (Liu and Fawcett 2008). However, many of the studies assessing pain are limited by small sample sizes, inclusion of different stages of cancer and sources of pain, or a lack of control conditions (Cassileth and Vickers 2004; Currin and Meister 2008; Ferrell-Torry and Glick 1993; Sturgeon et al. 2009). More research is needed to determine the long-term effects and the clinical relevance of massage as an adjunct for pain control.
The effect of massage on nausea is inconclusive, with some studies showing effectiveness and other studies showing no effect (table 17.1). Specific stimulation of the P6 acupressure point was more effective for nausea than body massage (Dibble et al. 2000; Dibble et al. 2007; Shin et al. 2004).
Although massage is standard care for lymphedema (Williams et al. 2002), other symptoms have not been studied sufficiently to draw conclusions. Patients report improved energy following massage, but few studies demonstrate effects on fatigue (table 17.1). In longitudinal studies, fatigue remained lower 6 weeks after 10 sessions of biweekly massage (Listing et al. 2009), but massage was not quite significantly better (p = 0.057) than control groups in reducing fatigue after four weekly massage sessions (Post-White et al. 2003).
Anecdotal and research evidence exist to support the use of massage for comfort, pleasure, and respite from the stress of cancer. Massage is generally accepted to promote relaxation, relieve muscle tension (Pruthi et al. 2009), and improve quality of life for patients receiving cancer treatment (Sturgeon et al. 2009) or at the end of life (Kutner et al. 2008; Smith et al. 2009). Patients, caregivers, and family members benefit from both giving and receiving massage (Field et al. 2001; Goodfellow 2003).
Early massage studies are limited by small sample sizes, high dropout rates, and lack of comparison groups or analysis between groups. Almost half of the early studies either lack a control group or fail to use randomization to groups. More recent studies in adults use randomized controlled trials (RCT) with larger sample sizes of 100 to 300 subjects (Campeau et al. 2007; Kutner et al. 2008; Mehling et al. 2007; Post-White et al. 2003). Considerable variation remains in the depth of touch and the types and dose of massage administered in individual studies, which makes comparisons among studies difficult and limits the translation of findings to practice.
Studies in Children With Cancer
Although massage is used clinically for children with cancer, only five studies have tested its effectiveness. Consistent with adult studies, decreased anxiety was the most commonly observed effect in children receiving massage for a variety of conditions (Beider, Mahrer, and Gold 2007) and for cancer (table 17.1). Other findings included improved mood (Field et al. 2001; Haun, Graham-Pole, and Shortley 2009), reduced discomfort (Haun, Graham-Pole, and Shortley 2009; Phipps et al. 2005), and decreased heart rate (Post-White, Fitzgerald, Savik, et al. 2009) and respiratory rate (Haun, Graham-Pole, and Shortley 2009). In these small studies, massage did not lower blood pressure (Haun, Graham-Pole, and Shortley 2009; Post-White, Fitzgerald, Savik, et al. 2009), lessen symptoms of pain, nausea, and fatigue, or reduce cortisol levels (Post-White, Fitzgerald, Savik, et al. 2009). Phipps and colleagues (2004) found that massage reduced the time to engraftment after bone marrow transplant, but had no effects on mood or symptoms. Importantly, these studies support the feasibility and safety of both massage for children with cancer provided by both therapists (Haun, Graham-Pole, and Shortley 2009; Phipps et al. 2004; Phipps et al. 2005; Post-White, Fitzgerald, Savik, et al. 2009) and parents (Field et al. 2001; Phipps et al. 2004; Phipps et al. 2005). In two studies on the effects of massage for parents of children with cancer, parents reported less anxiety (Post-White, Fitzgerald, Savik et al. 2009), less fatigue, and greater vigor (Iwasaki 2005).
Conducting massage research in children with cancer presents unique challenges. Small sample sizes and low statistical power are common problems (Beider and Moyer 2007; Phipps et al. 2004; Phipps et al. 2005), as are lack of standardization in dose, frequency, and style of massage (Underdown et al. 2006). Few self-reporting instruments are validated for children, and parent proxy reports are often an inaccurate reflection of the child’s perspective. Multisite studies make larger sample sizes possible, but they usually also require standardization of massage across settings, which may limit the effectiveness of treatment and the generalizability of study results. This is important because children often have unique needs and very specific tolerances and preferences that make standardization problematic. When children are undergoing treatment for cancer, massage therapists need to consider their health status and energy level, as well as their families’ overwhelming schedules. More than one massage session may be needed to help the child feel comfortable with therapist-provided massage.
Biological and Immunological Effects of Massage in Children and Adults With Cancer
The weakest evidence for massage effects is in biological and immunological outcomes. Some responses of the autonomic nervous system (heart rate, blood pressure) consistently decrease immediately following massage (Ahles et al. 1999; Billhult et al. 2009; Grealish, Lomasney, and Whiteman 2000; Post-White et al. 2003; Post-White, Fitzgerald, Savik, et al. 2009; Wilkie et al. 2000). However, effects on salivary cortisol and immunological measures of stress are often insignificant. Only Stringer, Swindell, and Dennis (2008) found decreased serum cortisol and prolactin 30 minutes after massage compared to a control group. No differences were captured beyond 30 minutes, suggesting a short-term response of this circulating stress-related hormone. Similarly, only one study by Billhult and colleagues (2009) showed a stabilizing effect of a massage on cytotoxicity of natural killer (NK) cells, with no change in numbers of NK cells.
Hernandez-Reif and colleagues (2004; 2005) provide the strongest evidence to date for biological and immunological effects in response to massage. Although sample sizes in one study were too small to compare immune effects by group (2005), increases in number of NK cells and positive psychological outcomes were associated with increases in dopamine and serotonin immediately after massage and again 5 weeks later. By contrast, Billhult and colleagues (2008) found no changes in oxytocin in patients with cancer, even though oxytocin has been shown to increase after repeated massage in other populations and in highly controlled studies using rat models (Lund et al. 2002; Wikstrom, Gunnarsson, and Nordin 2003).
Lack of evidence is not synonymous with lack of effect. Because of individual variability in immune responses and differential effects of acute and chronic stress, immunological effects are difficult to capture and to compare to normative values or standards. Detecting changes in response to massage requires large sample sizes, large effects, or both. Effects may also be missed because of an emphasis on single cross-sectional assessments or a measurement schedule that does not capture circulating responses to massage. Alternately, statistically significant effects can be found and erroneously attributed to massage if distributions are skewed or if other important variables (e.g., sleep, infection, medications) are uncontrolled. Statistical significance does not imply clinical significance; short-term immune changes captured on isolated posttest assessments may not have clinical relevance. Longitudinal studies will provide findings of greater clinical relevance.