Background

The global incidence of childhood cancer has observed a steady increase in the last decade likely due to increased access to treatment and improved reporting of childhood cancer [1]. Traditional and complementary medicine (T&CM) is a globally utilized supportive care tool in children undergoing treatment for malignancies in countries of all income levels [2, 3]; however, there is a general consensus that the evidence supporting its efficacy remains unclear for most indications. The lack of demonstrated safety and efficacy, the potential for adverse interactions with prescribed therapy, the delays in seeking conventional treatment, and the risk of diminishing the high cure rate obtained for several pediatric malignancies have raised concerns about T&CM use [4,5,6,7].

T&CM has the potential be a low-cost adjunct to existing supportive care regimens and may be especially useful in low-middle-income countries (LMICs) where consistent access to supportive care medications may be limited [8]. There is a precedent of utilizing T&CM as a low-cost approach for closing gaps in medical care in LMICs, particularly in rural areas [9,10,11,12]. It is evident that additional research in T&CM is necessary prior to its inclusion into supportive care regimens in pediatric oncology. To this end, we describe the results of a systematic review of clinical trials that investigated the efficacy of T&CM therapies for supportive care indications in childhood cancer.

Methods

Literature search

Our methodology followed the guidelines set forth by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [13]. Our search strategy included MeSH terms and text related to pediatrics, oncology, and T&CM (Online Resource 1). Our search was limited to studies of human subjects without any language restriction. All references were compiled into an EndNote (X7) library for review of titles and abstracts by two independent authors (AR and KT). Subsequent manual review of citations was performed with the inclusion of additional manuscripts that met the eligibility criteria below. Any disagreement was resolved by a final consensus (AR, KT, and EJL).

Eligibility criteria

Published manuscripts that reported on randomized, controlled, clinical trials that evaluated a T&CM therapy for supportive care purposes, performed among children and adolescents from birth to 18 years of age (inclusive), and diagnosed with a pediatric malignancy were included. Clinical trials that included adults were included in the systematic review if at least 80% of study participants were 18 years of age or younger. Classification of countries by income level was defined by criteria set forth by the World Bank [14]. Studies performed after cessation of cancer therapy, among children receiving surgery only, and case reports/case series/non-controlled trials were excluded. There was no exclusion by study date or date of publication.

Data extraction

Extracted data of interest included country of publication, year, demographic data (gender and age), diagnosis, study design, conventional cancer treatment, T&CM intervention (time in relation to phase of cancer therapy, dose, and duration of intervention), sample size, method of randomization, primary and secondary outcomes, statistical methods, and results. Data were extracted by one author (AR) and independently verified by a second author (EJL) using a standard data extraction sheet. Quality scores were calculated for eligible studies using the National Institute of Health’s Quality Assessment Tool for Controlled Intervention Studies, a 14-point scale that identifies the quality of randomized, controlled trials (Online Resource 2) [15]. The criteria for assessing study quality were adapted from previously published studies [16, 17]. Two reviewers (AR and EJL) extracted data for determination of study quality.

Data synthesis and analysis

Due to the heterogeneity of the data and a small number of clinical trials evaluating a single T&CM therapy, a formal statistical analysis was not feasible. Study descriptives were extracted and summarized in table format (Tables 1, 2, 3, and 4). Within each table, studies were further classified by study outcomes and by the income level of the country in which the study was performed.

Table 1 Summary of studies in acupuncture
Full size table
Table 2 Summary of studies in aromatherapy
Full size table
Table 3 Summary of studies in massage
Full size table
Table 4 Summary of studies in dietary supplements
Full size table

Results

Search strategy

A total of 6342 studies were produced in the original search (Fig. 1). Forty-seven papers from the original search met inclusion criteria. Of these, 16 studies were removed entirely due to inability to contact the author to clarify study details [18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33]. Thirty-one papers from the original search were included in the review. We identified an additional 56 manuscripts following the original literature search through reference review and ongoing monthly searches. Thirteen were eligible and included in the review [34,35,36,37,38,39,40,41,42,43,44,45,46]. Our final search results included 44 papers (acupuncture (N = 2), aromatherapy (N = 1), dietary supplements (N = 32), and massage (N = 9)).

Fig. 1
figure 1

Results of search strategy

Full size image

Acupuncture

Two studies reported on acupuncture, both investigating the effect on chemotherapy-induced nausea and vomiting (Table 1), one of poor [47] and one of good [48] quality [47, 48]. Both studies were performed in high-income countries (HIC) and with a small, heterogeneous population. Each study reported a significant decline in the use of antiemetics. One study reported a significant decrease in episodes of retching and/or vomiting [48].

Aromatherapy

One good-quality study investigated aromatherapy among children undergoing hematopoietic stem cell transplantation (HCT) (Table 2) [49]. This trial, performed in an HIC, examined the effects of bergamot essential oil on anxiety in 27 children undergoing HCT for a variety of diagnoses. The authors found increased nausea and anxiety in the aromatherapy group compared with the control group.

Massage

Nine studies, all performed in HIC and upper middle-income countries (UMIC), investigated the use of massage (Table 3) [39, 40, 50,51,52,53,54,55,56]. Six studies were of poor quality [40, 51, 53,54,55,56], and three were of fair quality [39, 50, 52]. Massage was administered in the inpatient [40, 50,51,52,53,54,55] and outpatient [54, 55] setting and at home [56]. One paper did not report the setting [39]. Various forms of massage therapy were provided and included massage therapy provided by parents [56], registered nurses [51], and licensed massage therapists [40, 50, 52,53,54,55]. One study did not report the provider of massage therapy [39]. Six of the studies examined the effect of massage on psychosocial outcomes [40, 52,53,54,55,56] and three studies on symptom management [39, 50, 51]. Three of the trials demonstrated a statistically significant reduction in child’s anxiety [40, 54, 56]. One study found that Swedish massage was effective at reducing nausea and vomiting during 48 h post chemotherapy (P = 0.027) [50], and another found that slow-stroke back massage reduced nausea severity and vomiting frequency over the course of six chemotherapy infusions [39]. A third found that massage therapy reduced pain [51]. Swedish massage provided in the inpatient and outpatient settings reported beneficial effects on muscle soreness, discomfort, respiratory rate, anxiety, emotional symptoms, and clinical progress scores [55].

Supplements

Thirty-two studies investigated the use of dietary supplements for several supportive care indications (Table 4). Twelve studies examined the effects of dietary supplements on mucositis, [34, 36, 37, 57,58,59,60,61,62,63,64,65], five studied treatment-related toxicities [46, 66,67,68,69], three examined appetite and weight management [35, 70, 71], three evaluated hepatic toxicity [41, 42, 72], three evaluated fever and neutropenia [38, 45, 73], two studies evaluated neuropathy [43, 44], two examined chemotherapy-induced nausea and vomiting [74, 75], and one study each examined bone mineral density [76] and gastrointestinal symptoms [77]. Of the 32 studies, 12 studies were performed in HIC [37, 43, 45, 57, 59, 61,62,63, 65, 72, 73, 76], 13 in UMIC [35, 46, 58, 60, 64, 66,67,68,69,70,71, 74, 77], and 7 in LMICs [34, 36, 38, 41, 42, 44, 75]. Seven papers received a quality score of “poor” [36, 46, 60, 66, 76, 77], 17 “fair” [34, 35, 38, 42,43,44,45, 58, 59, 62, 64, 67,68,69,70, 73, 74], and 9 “good” [37, 41, 57, 61, 63, 65, 71, 72, 75]. Most studies included a wide range of diagnoses with few studies performed among homogenous patient populations.

The use of dietary supplements for the prevention or treatment of mucositis was the most commonly investigated supportive care indication. Glutamine (N = 4) was the most widely studied supplement for this indication; however, variable doses, routes, and duration were studied (Table 4) [57, 60, 63, 64]. Two studies were performed in children undergoing HCT; one showed decreased use of morphine and TPN in children receiving glutamine [57], and the other showed no benefit [63]. The other two studies found decreased antibiotic use in the glutamine group [60], while the other reported no significant findings [64].

Three studies evaluated vitamin E for the prevention [65] and treatment [36, 58] of mucositis. One study found a significant improvement in mucositis scores [58]; the other two reported no significant findings [36, 65]. Vitamin A was evaluated for the prevention of mucositis and did not report significant results [59]. Finally, honey was found to reduce the recovery time of mucositis when compared to a mixture of honey, olive oil-propolis extract and beeswax, or control [34]. The use of propolis, a bee resin, alone did not produce any significant results [62]. The first T&CM clinical trial conducted through Children’s Oncology Group (COG) [61] administered Traumeel S or placebo to 200 children undergoing HCT. The authors did not find a significant effect on mucositis; however, a trend in the reduction in the administration of narcotics was observed.

Several studies examined T&CM therapies for a variety of treatment-related toxicities. Genistein [66] did not report significant effects, whereas beneficial effects were observed for selenium [46, 69]. One study examined Fuzheng Jianpi Decoction, a mixture of several different herbal remedies, and found improvement in anorexia, weakness, weight loss, constipation, pain, and somatic and psychological functioning [67]. Another study found a benefit on white blood cell (WBC) count and clinical symptom scores with various and individualized Chinese herbs [68].

Three studies addressed appetite and weight management [35, 70, 71]. Zinc chelate (2 mg/kg/day) significantly prevented weight loss, while also decreasing the number of infectious episodes [70]. An energy-dense eicosapentaenoic acid supplement (1 g BID) significantly decreased loss of body weight and body mass index [35]. A study evaluating PediaSure® and carnitine revealed no significant impact on anthropometric measures [71].

Three studies evaluated hepatic toxicity [41, 42, 72]. A small, multicenter pilot study found that milk thistle significantly decreased aspartate aminotransferase (AST) and total bilirubin among children with acute lymphoblastic leukemia in the maintenance phase of therapy [72]. Omega-3 fatty acids were found to reduce liver enzymes and increase antioxidants and uric acid [41]. Another study found that black seed oil decreased liver enzymes, alkaline phosphatase, and prothrombin time [42].

A study found that wheat germ extract significantly decreased neutropenic episodes and improved WBC and lymphocyte counts [73]. A Japanese study found that probiotics reduced the frequency and duration of febrile episodes and lowered the risk of developing fever [45]. In another study, administration of honey was associated with a reduction in the number of episodes of fever, number of children who developed febrile neutropenia, and reduced duration of hospital stays [38].

Glutamic acid was evaluated for neurotoxicity in two studies [43, 44]. One study found reduced severity of the tendon Achilles and patellar reflexes and decreased paresthesias, constipation, and neurotoxicity summary score [44]. In contrast, a multicenter consortium group study found that glutamic acid was not effective in the prevention of vincristine-induced neurotoxicity [43].

Two studies found improvement in nausea and vomiting with dietary supplements [74, 75]. A study examining vitamin A for D-xylose malabsorption found no significant effects [77]. Supplementation with calcitriol was found to improve lumbar spine bone mineral density [76].

Discussion

To the authors’ knowledge, we present the results from the first systematic review of clinical trials investigating T&CM interventions for supportive care indications in children and adolescents with cancer. Within each of the T&CM domains, the reported findings conflicted, identifying opportunities to further advance each of these domains within pediatric oncology. The widespread and persistent use of T&CM, particularly in LMICs, further endorses the need for additional research in pediatric oncology [3, 78, 79].

Several of the reviewed studies investigated the efficacy of massage therapy, a generally safe and accepted T&CM intervention [80]. We found encouraging evidence suggesting that massage therapy may be beneficial for several symptoms, which concurs with a recent consensus statement on non-pharmacologic approaches [81]. Evidence-based, non-pharmacologic T&CM interventions may be a cost-effective approach to advance the provision of supportive and palliative care across all income settings.

The role of acupuncture has been one of the most thoroughly researched T&CM modalities with some translational data describing its role for the treatment of several disorders [82,83,84,85,86], including chemotherapy-induced nausea and vomiting and pain management [87]. Our review found that there are a limited number of studies in pediatric oncology despite documented safety and feasibility in pediatric oncology [88,89,90]. Acupuncture may be especially beneficial for clinicians, children, and adolescents seeking non-pharmacologic approaches to manage a specified indication or symptom clusters. Training and licensing guidelines set forth by HIC or countries with an established system for delivering Traditional Chinese Medicine may serve as a framework for the investigation of acupuncture in countries without an established body of legislation.

Our review found that the largest number of T&CM studies evaluated the role of a dietary or herbal supplement for symptom management. The role of dietary supplements has been one of the most controversial aspects of T&CM due to the risk of adverse interactions with cancer therapy together with the absence of governing bodies providing oversight on the manufacturing and processing of dietary supplements. We found that two large cooperative groups conducted multicenter studies thus providing a framework for the conduct of T&CM. While several studies have reported encouraging results, the quality of the trials precludes their integration into existing standards of practice. We found that for select T&CM supplements, a benefit may be evident. This may have a significant impact in LMICs where access to supportive care medications may be scarce. In these settings, the risks and benefits of T&CM supplements should be weighed prior to their incorporation into care.

The strengths of our systematic review were the clearly defined eligibility criteria, the inclusion of a research librarian for the conduct of a systematic search, the evaluation of evidence from both HIC and LMIC, and the consideration of a quality score for each study. However, there are several limitations to our review, many of which are inherent to the conduct of systematic reviews. Several of the screened studies were not obtained due to inability to contact the authors or inability to locate the published article. It is also plausible that due to limited resources in LMICs, not all clinical studies were submitted for peer review publication. Therefore, we cannot exclude publication bias in our study. We were unable to conduct a formal meta-analysis due to the limited number of studies investigating the same indication and the heterogeneity within the studies that reported on the same outcomes. While we were able to identify areas that appear to be encouraging for future research, it must be recognized that our recommendations evolved from a limited number of clinical studies. Moreover, many of the studies received a low-quality score; thus, our findings are not based upon high-quality clinical trials. Finally, most of the included studies were performed in HIC, thus limiting the generalizability of their findings to the resources available and clinical care delivered to pediatric cancer units in LMICs.

There has been significant scientific effort in advancing the science of T&CM among children with cancer in both HIC and LMIC. Although most studies in this systematic review were of poor quality, a body of literature exists to foster educational and research initiatives. Pediatric cancer units interested in incorporating T&CM into the supportive care needs of children with cancer should consider the existing evidence alongside national policies, barriers in delivering existing care, and indigenous resources to identify the modalities that may be readily integrated into institutional clinical care and whose research findings will have an impact on the quality of care delivered by the institution.