Of these compounds, ellagic acid (EA) quickly became the phytochemical of research focus, demonstrating potent antioxidant and anticarcinogenic capabilities (Jurenka 2008). In vitro studies suggest an antiproliferative role for EA in a wide range of cancer cell lines. Neuroblastoma cell growth was inhibited by EA in concentrations between 10 and 30 μM (Fjaeraa and Nanberg 2009). As shown in Table 3, treatment with EA was also observed to significantly reduce cell growth in breast, prostate, colon, skin, ovarian, oral cavity, and bone cancer cells with IC₅₀ doses between 21.5 and 462 μM (6.5–139.6 μg/mL) (Han et al. 2006; Kasimsetty et al. 2010; Malik et al. 2011; Seeram et al. 2007).

In vivo research has demonstrated that a diet supplemented with EA in an estrogen-induced mammary tumor rat model significantly reduced the incidence of tumors by 19 %, tumor volume by 65 %, and tumor multiplicity by 41 % when compared to rats given a control diet (Aiyer and Gupta 2010). A mouse skin cancer model resulted in significant inhibition of DMBA-induced epidermal hyperplasia with topical application of EA while the incidence of aberrant crypt foci was significantly reduced in a DMH-induced colon cancer rat model (Kowalczyk et al. 2009; Umesalma and Sudhandiran 2010). Hamster buccal pouch carcinoma also induced by DMBA was nonexistent in hamsters with EA-supplemented diet, although hyperplasia was evident upon histopathological evaluation (Vidya Priyadarsini et al. 2012).

Cellular and animal investigations have thus presented strong evidence that EA is a promising phytochemical for the prevention and treatment of cancer. However, the doses required to provide these beneficial effects are considerably higher than the physiologically attainable blood plasma level of 0.018–0.033 μM (Mertens-Talcott et al. 2006; Seeram et al. 2004, 2006). Since inquiries into the anticancer properties of pomegranate phytochemicals other than EA have also shown potential for chemoprevention and treatment, the spotlight has turned to focus on properties of the whole pomegranate.

Confirmation of the anticancer benefits of whole pomegranate has been provided by a number of animal studies of prostate, skin, and lung cancer. Inhibition of proliferation by pomegranate extracts has been explored in several in vitro studies of prostate, lung, oral, and colon cell lines with 30–100 % inhibition of proliferation observed with doses from 10 to 150 μg/mL (Khan et al. 2007b; Malik et al. 2005; Seeram et al. 2005). Intriguingly, one of these studies determined that greater antiproliferative activity was provided by whole pomegranate juice when compared to a total pomegranate tannin extract, punicalagin, and ellagic acid at normalized concentrations of punicalagin (Seeram et al. 2005). Pomegranate juice metabolites, including ellagic acid and urolithin A, were found to be synergistically interacting among themselves, thus emphasizing the idea of phytocomplexity, to inhibit growth of PC-3 and DU145 cells, even causing death by affecting the cell cycle kinetics and induction of apoptosis (Vicinanza et al. 2013). Pomegranate extract was also found to inhibit proliferation of breast cancer cells while inducing apoptosis (Shirode et al. 2014). The authors attribute the antiproliferative effects to reduced DNA repair gene expression and induction of double-stranded DNA damage (Shirode et al. 2014). Pancreatic cancer cell growth was decreased with anIC₅₀value of 50 μg/mL while higher doses of 100 and 125 μg/mL were required for two oral cancer lines (Nair et al. 2011; Weisburg et al. 2010). In contrast, prostate cancer cells were inhibited by a whole pomegranate fruit extract (PFE) with an observed IC₅₀ value of only 5.7 μg/mL (Sartippour et al. 2008). In addition to antiproliferative activity, a protective role for PFE was suggested when UVA- and UVB-induced damage in normal human epithelial keratinocytes (NHEK) was inhibited by treatment with PFE (Afaq et al. 2005a; Syed et al. 2006). Follow-up studies in mice further illustrated the photochemopreventive potential of pomegranate with oral administration of PFE demonstrating inhibition of UVB-induced damage and early carcinogenesis (Afaq et al. 2010). A summary of these studies is found in Table  4 .

Prostate cancer was significantly inhibited in mouse xenograft models upon oral ingestion of PFE resulting in 35–60 % reduction in tumor volumes, delays in tumor incidence, and significantly increased survival times (Seeram et al. 2007; Malik et al. 2005; Adhami et al. 2012; Sartippour et al. 2008). Efficacy of PFE against lung cancer was explored in a xenograft model and a Benzo(a)pyrene [B(a)P] and N-nitroso-tris-chloroethylurea (NTCU)-induced model, both of which involved oral administration of PFE to mice which produced delayed progression of tumors, significantly increased survival time, and over 50 % reduction in tumor volume (Khan et al. 2007a, b). A 33 % tumor growth inhibition of tobacco smoke-induced lung adenomas was observed upon daily oral administration of pomegranate fruit extract (Naghma Khan et al. 2013). Skin tumors induced by 12-O-tetradecanoylphorbol-13-acetate (TPA) developed later, were fewer in number, and weighed less in mice treated with topical applications of PFE as opposed to control mice (Afaq et al. 2005b).

An initial human trial treating 46 men with rising prostate-specific antigen (PSA) values after primary treatment indicated a significant increase in the average PSA doubling time (PSADT) with daily oral ingestion of 8 oz of pomegranate juice. Untreated PSADT averaged 15 months while PSADT for treated patients averaged 54 months. As this was an open-label, single-arm, small clinical trial, future investigations are planned that will include more patients and placebo controls (Pantuck et al. 2006). More recently,a similar result of lengthened PSADT was observed in a randomized, double-blind phase II clinical trial, where PSA levels even declined in about 13 % of patients in the test group treated with pomegranate extract (POMx) (Paller et al. 2013). However, in direct contrast another placebo-controlled clinical trial in Switzerland re-reported that daily pomegranate intake does not influence the PSA levels when compared to the placebo group (Stenner-Liewen et al. 2013). It is important to note that these studies involved different populations, with the Swiss study treating advanced prostate cancer patients. Additionally, the studies used different preparations of pomegranate, a critical factor in this kind of research. The initial trial by Pantuck et al. used a juice made from crushing the whole fruit while the other studies used pomegranate extract capsules or a juice product that included 27 % pomegranate extract. As the quantity and composition of phytochemicals can vary widely based on preparation methods, the use of different study populations and products could be responsible for these contradictory results.

Several studies providing comparisons between whole products and fractions or constituents of pomegranate have suggested an additive or synergistic effect with treatment by whole pomegranate products. In pancreatic cancer cells, PFE treatment was observed to induce cell cycle arrest and inhibited cell proliferation at an IC₅₀ of 50 μg/mL, while a dose of 25–30 μM (7.55–9.07 μg/mL) produced the same results with ellagic acid. Given that ellagic acid comprises only 2.7 % of PFE, its contribution to the effects provided by whole pomegranate extract equates to only 1.35 μg/mL, indicating a minimum fivefold difference in the concentration of EA required when administered as an isolated compound (Nair et al. 2011). Additionally, a comparison between pomegranate extract and paclitaxel determined that PFE inhibited cell proliferation more quickly and more effectively than this standard chemotherapeutic drug (Nair et al. 2011).

Pomegranate juice (PJ) exhibited greater inhibition of proliferation than constituent compounds punicalagin, ellagic acid, and a standardized total pomegranate extract in a variety of cancer cell lines (Seeram et al. 2005). While all treatments demonstrated dose-dependent reductions in cell growth, the inhibition provided by PJ at very low doses was not achievable by other compounds without considerable increases in dosage. Similarly, a comparison of pomegranate juice (PJ), POMx (whole extract), punicalagin, and ellagic acid revealed superior antiproliferative activity of PJ in all three prostate cancer cell lines evaluated (Hong et al. 2008). It is important to note that both PJ and POMx were processed from the whole pomegranate and should therefore provide similar levels of inhibition, but the effects of POMx did not approach those of PJ until higher doses were utilized.

An interesting study evaluating antiproliferative activity of individual pomegranate phytochemicals compared to the whole extract found that while both the extract and the phytochemicals inhibited proliferation and reduced viability of mouse mammary cancer stem cells in a time- and dose-related manner, the doses of phytochemicals were higher than those found in the whole extract. EA inhibited cell growth with an IC₅₀ of ~7.5 μM (2.27 μg/mL), but its contribution to the IC₅₀ of PE was 1.35 μg/mL, a difference of 41 %, suggesting that the effect of EA is enhanced when administered as part of the whole extract (Dai et al. 2010).

Evaluation of the reported effective doses for whole pomegranate extracts versus ellagic acid provides considerable evidence for synergistic interactions between naturally occurring phytochemicals in pomegranates. Pomegranate extracts proved to be effective in doses between 50 and 200 μg/mL, providing an ellagic acid content between 0.15 and 5.4 μg/mL (Table  4 ). In contrast, the IC₅₀s for ellagic acid were reported to be between 22 and 462 μM, correlating to 6.5–139.6 μg/mL (Table 3), which are substantially higher, thus lending support to our idea that greater anticancer activity can be observed with whole foods than with constituents alone.

Much of the early tomato research focused on lycopene, a fat-soluble carotenoid providing the red pigment found in red tomatoes, watermelon, and other fruits (Kelkel et al. 2011). There is agreement that lycopene is a powerful antioxidant, often referred to as the most efficient carotenoid quencher of singlet oxygen (Di Mascio et al. 1989). In addition to antioxidant properties, lycopene displays potent anticancer activity including inhibition of cell growth, anti-inflammatory effects, cell cycle arrest, and alteration of cell signaling pathways (Kelkel et al. 2011). Interestingly, absorption of lycopene is affected by cooking as heat processing results in higher levels of cis-lycopene, the form found in human plasma, as opposed to the trans configuration dominant in fresh tomatoes (Tan et al. 2010). Raw tomatoes have an estimated lycopene concentration of 8.8–42 μg/g, whereas the range for tomato juice and tomato sauce is 86–100 and 63–131 μg/g, respectively (Russo et al. 2010). The bioavailability of lycopene can also be increased with the addition of lipids to the processing or ingestion of tomato products (Tan et al. 2010). However, there seems to be a maximum level of absorption regardless of the dose administered, suggesting the presence of saturable absorptive mechanisms (Diwadkar-Navsariwala et al. 2003). Keeping these factors in mind, cellular, animal, and human studies have provided considerable evidence for chemoprevention by lycopene.

Lycopene has been shown to significantly inhibit the proliferation of prostate, liver, lung, breast, and colon cancer cell lines with inhibition of at least 50 % cell growth for doses between 1.3 and 50 μM, concentrations that are well above the reported serum lycopene concentration of 0.34 and 0.65 μM achievable upon acute lycopene supplementation (Burgess et al. 2008; Diwadkar-Navsariwala et al. 2003; Hwang and Bowen 2005a, b; Hwang and Lee 2006; Koh et al. 2010; Salman et al. 2007; Tang et al. 2008; Palozza et al. 2010; Takeshima et al. 2014; Asmah Rahmat et al. 2002). Lycopene may exert more suppressive activity in androgen-independent prostate cancer cells as demonstrated by IC₅₀s of 26.6 and 40.3 μM compared to 168.5 μM for an androgen-dependent cell line (Tang et al. 2005). These studies are summarized in Table 5. However, several limitations exist in assessing the in vitro efficacy of tomato products containing lycopene as they are particularly sensitive to light, oxygen, and temperature in addition to their incompatibility as hydrophobic molecules in aqueous culture media (Ford et al. 2011; Hwang and Bowen 2005a).

Animal experiments have reinforced the link between lycopene and reduced risk of carcinogenesis. Androgen-independent xenograft prostate tumor growth rate in rats was significantly reduced from ~50 to 76 % by oral ingestion of lycopene with a correlating reduction in tumor volume of ~43–81 % for doses between 10 and 300 mg/kg (Tang et al. 2005). Oral consumption of lycopene has also been found to inhibit hepatoma tumor metastasis and reduce colon cancer tumor growth in nude mice models in addition to upregulating antioxidant and immune system activity in a gastric cancer rat model (Huang et al. 2008; Luo and Wu 2011; Tang et al. 2011). Human clinical studies of lycopene are under way and a number of these studies investigating lycopene’s purported anticancer benefits will likely provide valuable information regarding absorption, disposition, and any chemopreventive or therapeutic properties of lycopene.

The anticarcinogenic properties of tomatoes have also been attributed to flavonoids, which have individually displayed antiproliferative effects but when combined with other flavonoids the effects were far more striking, strongly suggestive of additive interactions between these phytochemicals (Campbell et al. 2004; 2006). Another line of research investigated the tomato glycoalkaloid referred to as tomatine, found in green tomatoes, with promising initial results (Friedman et al. 2009).

A 2007 review of earlier clinical trials concluded that while lycopene containing tomato products are significantly beneficial, the effects are not the result of lycopene exclusively, but rather the combination of phytochemicals and other components in the whole tomato (Basu and Imrhan 2007). Research on whole tomatoes and the phytocomplexity found naturally in tomatoes has provided support for this conclusion. As observed with other foods reviewed here, the results have suggested that superior benefits may be realized by utilizing the whole tomato (Fig. 3) as opposed to selectively using only one component.

Colon adenocarcinoma cell growth was inhibited and apoptosis induced in a dose-dependent manner by tomato digestate from in vitro simulated digestion of a tomato genetically engineered to be β-carotene rich (Palozza et al. 2009). A dose of 100 mL/L of tomato digestate inhibited cell growth by 55 %. This dose corresponds to a β-carotene concentration of 0.08 μM, but when purified β-carotene was compared in an identical assay, doses as high as 10 μM did not inhibit cell growth as effectively as tomato digestate, suggesting the involvement of other compounds or processes (Palozza et al. 2009). LycoC, a lycopene phytocomplex with 6 % purified lycopene, inhibited HL60 leukemia cells at concentrations between 10 and 25 μM which corresponds to a lycopene concentration of 0.6–1.5 μM (Ettorre et al. 2010). Explorations of extraction methods and cytotoxity in HT29 colon cancer cells determined the IC₅₀s of 62.5 μg/mL for the petroleum ether extract and 87.0 μg/mL for an in vitro digested tomato extract (Guil-Guerrero et al. 2011). In another study, Bhaumik et al. have performed cell growth inhibition studies employing a variety of tomato extracts prepared in various solvents including methanol, ethanol, acetone, and chloroform against HeLa cells. Results from this study revealed that these tomato extracts exhibited moderate efficacy with IC₅₀ values ranging from 57 to 69 μg/mL (Bhaumik 2014).

A comparison of the effects on human sera of ingestion of lycopene-rich red tomatoes (RTS), lycopene-free yellow tomatoes (YTS), and purified lycopene (LYS) confirmed similar lycopene concentrations in RTS and LYS sera, but produced some surprising results in terms of anticarcinogenic gene expression (Talvas et al. 2010). While increases in beneficial gene expression were observed in RTS and YTS, decreases of anticarcinogenic and increases of pro-carcinogenic gene expression were revealed with evaluation of LYS sera samples, suggesting that the food matrix in whole tomatoes may provide positive benefits and modulate negative effects. While cell proliferation was unaffected by any of the samples at physiological levels, it may be important to note that cells in this experiment were exposed to the human subject sera for only 24 h while a number of other studies have utilized 48 h or longer exposures (Talvas et al. 2010).

Not all investigations into the anticancer properties of tomatoes have focused on lycopene. Research of the isolated green tomato glycoalkaloid tomatine revealed growth inhibition of stomach, liver, colon, and breast cancer cell lines when treated with purified α-tomatine. Inhibition was also observed in normal liver cells. When extracts of a variety of green tomatoes were tested, a number of them were also found to inhibit these cell lines. A comparison of the IC₅₀s for purified tomatine and the quantities of tomatine in the green tomato extracts suggests that the blend of glycoalkaloids found in these tomatoes is more effective than the individual component. Results for HepG2 liver cancer cells and MCF-7 breast cancer cells are shown in Tables 5 and 6.