Ethnobotanical Study of Medicinal Plants Used by Traditional Healers to Treat Cancer-Like Symptoms in Eleven Districts, Ethiopia

There is no ethnobotanical study conducted specifically on medicinal plants traditionally used to treat cancer in Ethiopia. Yet, traditional herbalists in different parts of the country claim that they have been treating cancer-like symptoms using herbal remedies. The objective of this study was to document medicinal plants traditionally used to treat cancer-like symptoms in eleven districts, Ethiopia. Traditional herbalists were interviewed using semistructured questionnaires, and field visits were also carried out to collect claimed plants for identification purpose. Seventy-four traditional herbalists, who claimed that they knew about and/or had used medicinal plants to treat cancer-like symptoms, were selected using the snowball method and interviewed. Herbalists used their intuition and relied on the chronicity, growth of external mass, and spreading of the disease to other parts of the body, as a means to characterize cancer symptoms. Furthermore, in some of the study districts, herbalists reported that they treat patients who had already been diagnosed in modern healthcare institutions prior to seeking help from them. The inventory of medicinal plants is summarized in a synoptic table, which contains the scientific and vernacular names of the plants, their geographical location, the parts of the plants, and the methods used to prepare the remedies. A total of 53 traditionally used anticancer plants, belonging to 30 families, were identified during the survey. The most frequently reported anticancer plants were Acmella caulirhiza Del (Asteraceae), Clematis simensis Fresen. (Ranunculaceae), Croton macrostachyus Del. (Euphorbiaceae), and Dorstenia barnimiana Schweinf. (Moraceae). Organizing traditional healers, documenting their indigenous knowledge, and scientifically validating it for the development of better cancer therapeutic agents constitute an urgent and important task for policymakers and scientists.


Introduction
Cancer is a complex disease that is very heterogenic and variable at cellular level and also differs from one patient to the other in its behaviour, development, and outcome [1]. Physical, metabolic, and behavioural variations of cancer cells from normal ones arise through the accumulation of genetic modifications and help them to proliferate rapidly, escape from host immune surveillance, and ultimately invade distant tissues [2]. Histopathological, genetic, and epigenetic and clinical outcome variations between and within different types of cancers have been the greatest challenge to understand the disease and develop novel therapies [3].
Surgery and radiation therapy were the most preferred means of treatment to control cancer before 1950 and after 1960, respectively [4]. Chemotherapy can be done before surgery to shrink the tumor or after surgery to kill the remaining cancer cells [5]. However, most of the chemotherapeutic drugs lack specificity and tend to rapidly damage normal dividing tissues, causing side effects such as immunosuppression, neurotoxicity, and hair loss [6]. Moreover, resistance has also reduced therapeutic efficacy of some anticancer chemotherapeutic drugs [7].
In order to address these limitations, tapping nature as a major source of chemically diverse novel anticancer compounds is a consistently proven track [8]. Screening natural products yield more hit with more "drug-like" characteristics (absorption and metabolism) as compared to screening of rationally designed compounds [9]. Furthermore, screening medicinal plants based on traditional use provides a higher chance of finding active plants relative to the random approach [10].
Ethiopia has a rich and diverse heritage of traditional medical practices, known for using plants to prepare more than 90% of the remedies [11]. In addition, the country has more than 6,500 higher plant species of which, around 12% are endemic [12]. Reports indicate that up to 80% of the population relies on traditional remedies as a primary source of health care [13]. Only few ethnobotanical reports from different agroecological zones of Ethiopia are available in the literature regarding medicinal plants used for cancer treatment.
In view of this fact and considering the weak traditional recording and knowledge transfer system and an alarming rate of environmental degradation, finding anticancer plants and documenting their ethnobotanical information constitute an urgent and indispensable task. erefore, the main aim of this study was to establish an inventory of medicinal plants traditionally used to treat cancer in eleven districts of Ethiopia.

Description of the Study Areas.
is ethnobotanical study was conducted in four national regional states of Ethiopia: Oromia, Amhara, Afar, and Southern Nations, Nationalities, and People. e survey included different districts from each region, namely, Bale Robe and Goba from Oromia, Bahir Dar Zuria and Filiklik from Amhara, Gewane from Afar, and Wondo Genet, Sodo Zuria, Doyo Gena, North Bench, Mizan Aman, and Shako from Southern Nations, Nationalities, and People Regional State (Figure 1). ese geographically, culturally, and agroecologically different study areas (Table 1) were selected mainly based on the availability of traditional healers and recommendations from health workers.

Data Collection.
A team comprising a botanist and researchers from Addis Ababa University was set up, and health authorities were contacted for permission and identification of traditional herbalists living in each study area. Altogether, 117 traditional healers were approached using the snowball technique and 74 traditional healers who used herbs to manage cancer-like symptoms were selected. Ethnobotanical data were collected between January and August 2016, mainly through individual interviews with the selected traditional herbalists using a semistructured interview questionnaire. e questionnaire was prepared in Amharic language and translated to different local languages for traditional healers who do not speak Amharic. is questionnaire was designed to obtain information in the following areas: (i) general data on the informant, (ii) school attendance, (iii) use of plants for cancer treatment, (iv) source of the plant material, (v) part of the plant used, (vi) method of medicinal preparation, (vii) route of administration, and (viii) side effects.
A traditional healer for the purpose of this study is "a person who is recognized by the community in which s/he lives as competent to provide healthcare by using plants and plant products." Each traditional healer was approached, briefed about the purpose of the research, and asked for his/ her verbal consent in talking about cancer and its treatment. ey were assured of the confidentiality of the information they provided. If plants were mentioned for their anticancer purposes, a botanical sample was collected. ese specimens were pressed and preserved for later identification at the National Herbarium, Addis Ababa University, Addis Ababa, and a voucher specimen of each plant was deposited in the institute. All botanical names have been transcribed according to the nomenclature system used by the Plant List (http://www.theplantlist.org).

Data Analysis.
e relative importance of medicinal plants used in the management of cancer-like symptoms in study areas was assessed using the relative frequency of citation (RFC), use value (UV), informants consensus factor (ICF), and cultural importance index (CI).

Relative Frequency of Citation (RFC).
e RFC was calculated by dividing the number of informants that cite a particular plant species (FC) by the total number of informants in the survey (N) [29]:

Use Value (UV).
e use value demonstrates the relative importance of plant species to treat particular ailment, and it is determined by the following formula [30]: where "UV" stands for the use value of a species, "U i " stands for the number of use reports cited by informants for that plant species, and "N i " is the total number of informers who reported the particular plant species i.
where Nur is the number of use reports of informants for particular ailment category and Nt refers to the number of species used for the ailment category by all informants.

Cultural Importance Index (CI).
Cultural importance index (CI) is calculated by the sum of the use reports (UR) of informants mentioning each species use (from i 1 to i N ) in each use category and adding all the UR of each category (from u 1 to u NC ) divided by the total number of informants N. is index is determined by the following formula [29]: where CI is an ethnobotanical index that indicates the spread of the use along with the diversity of uses of each species.

Results
e informants consisted of 66 male and 8 female traditional healers and they were divided into three age groups: 20-40, 41-60, and ≥ 61 years. Out of 74 interviewed traditional healers, most of them (N%) were adults aged between 41 and 60 years. Majority of the respondents (70.2%) gained their knowledge from family members and 82% of all interviewed respondents practiced ethnomedicine for more than 25 years. More than 70% of the respondents were either only at their primary level of education or did not have a formal education at all (Figure 2). Traditional healers usually used their intuition and relied on the chronicity and growth of external mass, as a means to diagnose cancer. Lumpy growth was the most commonly cited criteria used to diagnose cancer, followed by ulcerative wounds and bleeding (Table 2). However, there were instances where some of the healers claimed to have treated patients already diagnosed with cancer at modern health institutions. Traditional  Evidence-Based Complementary and Alternative Medicine healers identified cancer as "Nekersa" in Bahir Dar Zuria and Filiklik, "Naqarsa" in Bale Robe and Goba, "Sissac" in Gewane, "Xoka or Toka" in Doyo Gena, "Balamo" in Wondo Genet, "Kums or niamt" in North Bench, and "Kanser" in Sheko and Sodo Zuria district. Out of the 6 specific cancer types (skin, breast, lung, cervical, throat, and intestinal) claimed to be treated by the respondents, skin cancer was a dominant one followed by breast cancer.
A total of 53 plant species belonging to 30 families were reported for their anticancer use (Table 3). e result of this study showed that shrubs (49.1%), herbs (33.9%), trees (13.2%), and climbers (3.8%) were the main sources of potential anticancer medicinal plants. is study also indicated that leaves (56.7%) were the most commonly used plant parts followed by roots (21.7%), bark (6.7%), stem (1.7%), seeds (1.7%), whole plant (1.7%), leaves and roots (5%), leaves or stem (1.7%), and leaves or seeds (1.7%) ( Figure 3). Most of the reported plants occurred naturally in wild (96.2%); however, cultivation was also a source (3.8%). Reported medicinal plants have been traditionally claimed to be used to treat different types of ailments including cancer. However, only few have been scientifically investigated for their antiproliferative or cytotoxic activity (Table 4). While comparing the amount and distribution of anticancer plants in the past ten years, regardless of the study areas, all respondents believed that the amount and distribution of these plants are reduced.
In the current study, the highest UVs were recorded for Aloe spp. (6), Albizia schimperiana (4), Sida schimperiana (4), Achyranthes aspera (4), Brucea antidysenterica (4), Cleome brachycarpa (3), Leonotis ocymifolia (3), and Prunus africana (3). e lowest UVs were obtained for Acokanthera schimperi, Acmella caulirhiza, Cineraria abyssinica, and Gnidia involucrata (Table 3). A total of 228 use reports have been documented and categorized into seven categories (Table 5). Among these, other ailments (46.3%) and skin cancer (26.5%) had the highest use reports. Furthermore, ICF values were also calculated and ranged from 0 to 0.42. e highest ICF values were recorded in other ailments (0.42) and breast cancer (0.32) followed by skin cancer (0.23) category (Table 5). e other ailments category comprises of diseases such as stomach ache, malaria, wart, swelling, wounds, evil eye, toothache, bleeding, gastrointestinal disorder, headache, bone fracture, cough, snake bite, herpes simplex, tonsillitis, hypertension, dandruff, fever, and hemorrhoid. e ICF value of the remaining four categories (lung cancer, colon cancer, cervical cancer, and throat cancer) was zero. Quantitative ethnobotanical indexes such as RFC and CI were calculated in this study to shows that the Pearson correlation coefficient of RFC was positively and negatively correlated to CI and UV, respectively ( Table 6). Most of the reported remedies, prepared from these plants, were either applied topically (50%) or taken orally (41.7%). e remaining remedies were prepared to be administered either topically or orally (3.3%), both topically and orally (1.7%), and intranasally (1.7%). Usually, fresh plants were finely chopped, dried, and pounded to powder form. en, the powder of either one or the combination of more than one plant was either mixed with drinking water or pasted and applied topically. In other cases, fresh plant parts were decocted and taken orally or crushed and applied topically. Water was the main medium in preparation of most remedies and additives like honey, milk, and butter were also used. To determine the amount of plant parts used to prepare remedies, traditional healers used spoon, fingertip, and number (in case of fresh leaves). Adverse effects reported by respondents include vomiting, diarrhea, and skin ulcers.

Discussion
Despite the rich biodiversity of the study areas, broad acceptability, and centuries-old tradition of using traditional medicines, the number of anticancer plants reported in this study is far less than expected. As it was reported by different ethnobotanical studies conducted in different parts of Ethiopia, this could be attributed to the attitude of many traditional healers to guard their indigenous medical knowledge as a family secret and hence hesitant to share with the researchers [13,32,73]. Justifying the lower number of female traditional healers (8, 11%) participated in this study, these studies also inferred that traditional healers usually pass their knowledge to the first son of the family.
No previous reports 16 Evidence-Based Complementary and Alternative Medicine of the study sites and the difficulty to pinpoint the physical symptoms of lung, colon, cervical, and throat cancer as compared to the breast and skin cancer. e present study also revealed that RFC and CI values of some reported species are similar. However, there is a distinct difference in species ranking using each index. C. macrostachyus is placed in the first position according to both RFC and CI index. is could be due to the fact that this species is mentioned by many informants and is the most recognized plant in most study areas. Furthermore, CI value of C. macrostachyus is also high, suggesting the diversified use of the plant. V. auriculifera and C. simensis ranked next to C. macrostachyus, according to RFC index. On the other hand, D. barnimiana and Aloe spp. ranked 2 nd and 3 rd by CI index. It has been suggested that UV value is a good measure of use diversity, than the number of citations [144]. In agreement with this, UV value in our study is driven by species with greatest number of use rather than those cited by more informants. e Pearson correlation coefficient of −0.36, between RFC and UV, shows significant negative association between the local importance of each medicinal plant and relative importance of use of plants. is result is in contrast to previous studies that reported a significant positive correlation between RFC and UV [145,146]. On the other hand, there is a significant positive correlation between RFC and CI (r 2 � 0.74, p < 0.001) implying that their pattern matches across species. e species with larger RFC value usually have higher CI, such as Croton macrostachyus and Vernonia auriculifera.
Leaves and roots are the most commonly used plant parts in the preparation of remedies in the study districts.
Similarly, other ethnobotanical studies conducted in different parts of Ethiopia also reported that leaves are the dominant plant part followed by root [16][17][18]20]. e preference towards leaves may be because leaves are the main photosynthetic organs in plants and the primary reservoirs for secondary metabolites with medicinal values [36]. In contrary to other ethnobotanical studies [17,18], where the common use of concoctions and oral route were reported, in the current study majority of the reported remedies are prepared from a single plant species and applied topically.
Although herbal remedies are believed by the general public to be safe [46], some research findings suggested otherwise. For instance, traditionally used ai anticancer plants Ganoderma lucidum (Fr.) Karst., Houttuynia cordata unb., and Saussurea involucrata Matsum. & Koidz. were reported to cause side effects such as headache, insomnia, constipation, and diarrhea [62]. Similarly, side effects such as vomiting, diarrhea, and skin necrosis, associated with the use of traditional herbal remedies, were reported in this and other ethnobotanical studies conducted in Ethiopia [149,150]. Few side effects reported in this study, as compared to other ethnobotanical studies conducted in Ethiopia, could be attributed to the frequent use of the topical route of administration. Nevertheless, considering the probability of underreporting adverse effects, extensive toxicological investigations should be conducted to protect the public.
In vitro cytotoxicity and antioxidant properties of some of the plants reported in our study have also been studied. Among these plants, potent cytotoxic activity was reported for knipholone anthrone isolated from Kniphofia foliosa, with IC 50 value that ranges between 0.9 ± 0.1 and 3.3 ± 0.4 μg/mL [89]. Similarly, Nibret and Wink reported the cytotoxic activity of the crude extract of Acokanthera  schimperi with IC 50 value of 7.1 μg/mL [73]. Studies conducted on the leaves of Cineraria abyssinica [100], bark of Senna singueana [116], and bark and leaves of Rumex nepalensis [79] also revealed potent radical scavenging activity of these plants.

Conclusion
e present study showed that traditional healers in eleven districts of Ethiopia use different medicinal plants to manage cancer-like symptoms. Frequency of citation value ranked Croton macrostachyus Del., Clematis simensis Fresen., Dorstenia barnimiana Schweinf, Vernonia auriculifera Hiern, and Acmella caulirhiza Del. as most cited plant species in study areas. Hence, based on these findings, we are currently evaluating the in vitro antiproliferative activities of reported medicinal plant species with a higher frequency of citation against human breast adenocarcinoma (MCF-7), human uterine cervical adenocarcinoma (SiSo), human lung carcinoma (A-427), and human bladder cancer (RT-4) cell lines using crystal violate assay. However, considering the rapid disappearance of the traditional knowledge of medicinal plants and an urgent need for new anticancer agents, additional studies have to be conducted to document and scientifically validate traditionally used Ethiopian anticancer plants.
Data Availability e authors declare that all data supporting the finding of this study are included in this article and its supplementary information files.

Ethical Approval
Ethical approval was obtained from Addis Ababa University, College of Health Sciences Ethics Review Board (Ref no. ERB/SOP/126/12/2015).

Consent
Each participant consented before the interview.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
A.B., E.E., and K.A. jointly conceived the study. S.T. conducted the ethnobotanical study and taxonomical identification. S.T., A.B., E.E., T.G., and K.A. enriched the draft manuscript for its intellectual content. All authors read and approved the final manuscript.