Is Exposure to Titanium Dioxide Nanoparticle Associated with Occupational Lung Cancer among Titanium Dioxide Production Workers? An Emerging Issue ()
1. Introduction
Richard Feynman a University Professor in 1959 proposed the idea of nanotechnology engineering [1] . “There’s plenty of room at the bottom” was Richard’s famous quote relating to the concept of working with nanoscopic materials of sizes ranging from 0.1 - 100 nm [2] . Nanotechnology is a rapidly developing industry throughout the world since the idea proposed by Richard. This engineering discovery has brought about medical, engineering, and economic prosperity to most industrialised countries. For example, some industrial paints, food, computer chips, and sunscreens are manufactured from nanomaterial [3] . It is the physicochemical properties of nanomaterial that makes them very economical because of their large surface area, structure, solubility, super strength than steel and an excellent conductor of electricity [4] .
The scientific triumph has also brought with it emerging health concerns according to most literatures. This is a concern within occupational health setting where manufacturers involved in the production of nano engineered products may be exposed to nano dusts and particles [5] . In addition, nanomaterial is widespread throughout the world presenting new challenges to risk assessment managers to find solutions to control risks and hazards associated with the use of Carbon Nano Tubes and nanomaterial.
This review is focused on Titanium dioxide nano particle that is widely used to manufacture colour in paint products, plastics, papers, inks, food colour, toothpaste, and skin products [6] . The nano particle has been known to cause inflammation through entering a biological system through dermal; ingestion; and inhalation pathway. Moreover, Titanium dioxide nanoparticle is known to cause lung inflammation and the particle can also redistribute from the lung to kidney, liver, and brain and it is closely associated with ROS (Reactive Oxygen Species) [7] . Moreover, Titanium dioxide nanoparticle has been shown to induce ROS in mitochondria leading to DNA damage in lung cells.
Titanium dioxide Nano particle comes in three forms which include Rutile (mineral form), Anatase (Mineral and crystal form); and Brookite. Anatase form of TiO2 NP is the most reactive form known to cause inflammation to lung of laboratory rats than Rutile and Brookite, however, TiO2 NP normally exist in both forms in general [8] .
2. Methods
The research question for this systematic literature review was constructed using Cochrane “PICO” method. Since there is lack of long term epidemiological study design using cohort based studies from human exposure it is difficult to formulate research question using intervention type questions. Therefore, only “P” and “O” from PICO were used to construct the question. That is to design the review question using Population that is affected from exposure to a disease as the outcome. This approach is recommended by Cochrane Library Tutorial.
2.1. Literature Search Strategy
After the research question was formulated a search was conducted using QUT Library electronic data base. The search was conducted on the 11, April 2014 using the quick search box as shown in the 1stbox of Figure 1. The 2nd lot of search was done on the 14th of April 2014 using the health data base and narrowing the search to Medline and Medline Web of knowledge. The same search engine was used for the 3rd link which led to EBSCO host as shown in the 3rd box of Figure 1.
2.2. Key Search Terms
The keywords used in this order include Titanium dioxide nanoparticles lung cancer toxicity. In the first quick search button the following strategy was used
![]()
Figure 1. Shows how the Systematic Literature search was done.
as noted, “Titanium dioxide nanoparticle” AND “Lung cancer” was entered resulting in 1904 papers as indicated in Figure 1.
As with the other two data base searches the keywords were entered without open and closing inverted comas resulting in 149 papers for Medline Web of knowledge link and 14 papers for EBSCO host.
2.3. Refined Search
The search was refined to extract only research papers that has the following attributes which include only full text papers, journal format, available online, peer reviewed, and only in English language, and papers published within 2004-2014. The reason for choosing articles within this period is to gain the latest information on Titanium dioxide nanoparticle toxicity. The articles that were excluded are papers that do not have key word search terms, newspapers, reviews and letters. It is also worth noting that the contents of the papers were screened and those not related to the research question were eliminated. Moreover, only primary sources were included while secondary sources were not considered. The search results from the refined strategy yielded a sum of 286 articles from the three search data bases.
2.4. CASP (Critical Appraisal Skills Programme) Method
CASP [9] was employed to scrutinise the remaining 286 articles to obtain papers that met the criteria for the research question. The reason for further scrutiny is to ensure that only quality research papers were extracted, being valid, with results being statistically significant, and the outcome of the papers did answer the research question to prevent biases.
Refer to Figure 1 for explanation on methods used to select articles for the review.
2.4.1. Filtering Questions
The following questions were used to extract quality papers from the 286 papers. The values assigned to each question asked are;
YES = 1, Unsure = 0.5, and No relevance = 0
1. IS IT A TRUST WORTHY STUDY?
1) Was the objective of the study explicitly explained?
2) Were indicators/interventions explicitly described?
3) Is the research design relevant for human toxicity studies?
4) Is the sample size and sample reflective of the population?
5) Are confounding variables stated in the study that may affect the results?
6) Was pathogenesis described clearly in the study?
7) Was aetiology and prognosis clearly described in the study?
8) Was objective measurement conducted to minimise bias?
2. IS THE RESULT OF THE STUDY TRUST WORTHY?
9) Are the tests results reproduce able in the next experiment?
10) Are the results statistically significant?
3. IS THE STUDY OUTCOME BENEFICIAL TO THE SOCIETY?
11) Are the test results relevant to human exposure?
12) Will the results contribute to new knowledge of nanoparticle toxicity studies?
13) Will the results assist in mitigating nanoparticle hazard and risks within workplace setting?
14) Are the results conclusive?
2.4.2. Assessing Quality Papers Based on CASP Method
Table 1 below shows how quality papers were selected following the CASP method guided by the above fourteen (14) questions with ratings of 0 to 1; 1 representing YES, there is relevance, 0.5 being for UNSURE and 0 for NO RELEVANCE to the study.
2.4.3. Author’s Total Scores Based on the Fourteen Filtering Questions Asked
As shown in Figure 2 three (N = 3) of the papers reviewed had the highest score of fourteen (14) out of 14 questions, four (N = 4) of the papers had a higher score of 12 - 13 out of 14 questions, three (N = 3) papers had a high score of 11 - 11.5 out of14 questions meeting the quality screening test using the CASP method to be selected for our review. Only one (N = 1) paper with a score of 7.5 was selected for the review because of interesting results.
3. Results
A total of eleven (N = 11) articles were selected for the review based on the paper
![]()
Table 1. Shows Ratings for each paper based on 14 filtering questions.
As shown in Table 1 Eleven (N = 11) papers were screened based on the fourteen filtering questions asked and were each given a score of 0 to 1 to determine their overall score.
selection criteria as indicated.
As shown in Table 2 results from both cohort studies on occupationally exposed workers to titanium dioxide workers revealed no significant risk from carcinogenic induced respiratory diseases. The mathematical modelling based study
![]()
![]()
Table 2. Showing summary of the findings in the eleven (11) studies.
provided uncertainties about the association between TiO2 and respiratory induced malignant outcomes. While results from the 8 experimental studies done on rats showed no significant toxic risk from exposure to anatase nanoparticles while the results yield from the mice studies showed TiO2 did cause cell inflammation with minimal lung toxicity. Interestingly, in the human cells studied there was positive association between TiO2 with cytotoxicity and genotoxicity except the study by Belade et al. 2012.
4. Discussion
From the 11 papers that were extracted after the quality assessment the paper by Ling et al., 2011 was omitted for analysis because the study was based on mathematical modelling extracted from other studies. The discussion will be based on Table 2 page 7 - 9 including the summarized data based on the CASP screening questions on Table 1 p. 6 and Figure 2 p. 7.
4.1. General Analysis
Out of the ten (N = 10) studies, two (N = 2) were cohort design studies while eight (N = 8) were experimental based studies. The two cohort studies were long term studies which required a follow up on the same subjects over a period of time after the initial study. This was necessary to establish if there was an association of mortality cases among TiO2 processing workers exposed to TiO2 dusts. The eight (8) experimental studies on the other hand employed biological modelling to make inferences based on their objectives. The experimental designs involve an array of tests to measure outcomes in cytotoxicity, genotoxicity, enzymatic activity, cell apoptosis, inflammation, and cellular oxidative phosphorylation as the main examples documented.
In general, the 10 studies presented conflicting conclusions and this is because of the various research designs, objectives, methods, type of sample chosen, sample size, and the type of end points the researchers were contemplating to achieve. The end points can include short term studies that relate to acute toxicity testing or long term studies involving chronicity testing for carcinogenicity
![]()
Figure 2. Shows the Authors and their score based on theCASP 14 filtering questions.
tests. Most of the studies lacked properly designed research question, refer to Table 2.
A number of experimental studies developed more than one experiment to assess a certain phenomenon of interest like cytotoxicity, genotoxicity, and oxidative phosphorylation [10] [11] [12] [13] [14] . This was the case with Osman et al., 2010 and Srivastava et al., 2012. It was a challenging task to discuss each of these experiments therefore the discussion in this review will relate to the review question of these papers. In addition, the strengths and limitations of each paper will be discussed starting with the cohort studies by Boffeta et al., 2004, Ellis et al., 2010, and the experimental studies. The review will conclude with the limitations and concluding remarks.
4.2. Observational and Cohort Study Design
The two cohort studies by Boffeta et al., 2004, and Ellis et al., 2010 are generally of high quality with inclusion of large sample sizes, sound methodology, and clear outcomes.
Cohort study design minimises selection bias as it is generally a long term study on certain groups of subjects that require follow up. Statistical precision of such studies will generally be significant having a P value of <0.05 or much higher example P<0.001 to determine the association between the exposure or non-exposure scenarios. In the context of this review the authors aim to ascertain the association between exposure to TiO2 Nanoparticle and Lung cancer in TiO2 processing plants workers. The two selected studies are relevant for this review as they clearly demonstrated in their outcomes that there was no association between malignant and non-malignant respiratory diseases. A narrative approach is taken to discuss the research results other than the use of statistical software which is a limitation of this review.
The studies by Ellis et al., 2010 was conducted for a 29-year period studying workers exposed to TiO2 process within three DuPont Plants in the USA. Ellis et al., 2010 looked at employee’s exposure to TiO2 between 1935 when the three Plants were constructed and in operation until 2006. It was not clearly established how the follow ups were done, however, it was documented by the authors that they obtained the records from those Plants where their subjects were employed. These involved gathering updated demographic records such as employment history, death certificates, record of termination from employment, age, and gender as some examples [15] . As for Boffetta et al., 2004 their study looked at cohorts between 1959-1972 and 1997-2001.
As a result of the great length of follow up the outcome of their studies were conclusive. Interestingly, both studies presented similar results because their study design and sample size were similar. Furthermore, both authors admitted to confounding variables particularly noting smoking history of the workers with few cases of employees exposed to asbestos. Despite the confounding variables the outcome of the two studies showed no association between respiratory related diseases and TiO2 dusts in processing Plants [16] . Although the results appear significant there could be room for selection bias, in terms of recruitment and responses elicited. For instance, Ellis et al., 2010 admitted that they assumed all employees working in one of the three DuPont Plants to be Caucasians when there were no office records presented to them on the ethnicity of the subjects.
4.3. Experimental Study
In this study a total of 8 laboratory based experiments were done using biological systems in a controlled environment.
The biological model involved the use of rat cells, human cells, and whole animal body exposure for which certain toxicity studies can be done. The common toxicity studies include a control and a treated sample that were manipulated to achieve an outcome by the manipulator. Animal and human cells were exposed to TiO2 nanoparticle at various concentrations, and exposure duration, with observance on endpoints. The methods employed for the tests involved in vitro, in vivo, and whole organism exposure tests. All the studies generated various conclusions and this depended on their study objectives, study design, and methodology that were used.
Experimental studies do have strengths and limitations just like any other study design. A key advantage is that Scientists can draw quick conclusions from causal inferences and incidence rates and several outcomes can be assessed in which similar experiments can be repeated. The sample size for the 8 toxicity studies were not clearly specified. It appears the sample sizes used were small and the results demonstrated as statistically significant by the authors can be false due to type1 error. For instance, a study by Ambalavanan et al., 2014 only used 6 mice for the test and concluded that single dose of TiO2 caused inflammation to the rat pubs. This is similar to the experiment by Grassian et al., 2007 who used 6x 6 week old mice for whole body exposure. As one of the key limitations of experimental studies, generalizability of the results may have been done hence showing the results to be significant [17] .
Moreover, due to stringent ethical conditions scientific experiments can be easily conducted using other methods such as in vivo and in vitro methods using bio assays. This is necessary to extrapolate results in the experimental model to make inferences to human subjects. In this context the cell toxicity results in TiO2 exposure can be extrapolated to human exposure situations within workplace settings. The limitations of the 8 experiments with reference to the review question of this paper are that the sample sizes used were not well defined and appear to be small which could have led to generalization of results. Moreover, Titanium dioxide nanoparticle is already confirmed by IARC as group 2B carcinogen due to toxicity evidence from experimental studies. This is a signal for appropriate Agencies throughout the world to utilize the available data and be proactive in instigating control measures to prevent similar experiences with Asbestos like cancer as a point of reference.
4.4. Limitations
This review has limitations due to the following points. Firstly, it has not searched other data bases around the world to provide adequate data for the review. If other data bases were explored the results presented here may be different.
Secondly, only articles in English language were chosen restricting other articles that may present important findings in other languages. Thirdly, the experimental model results may not truly reflect the real exposure scenario of worker’s exposure to TiO2 nanoparticles in the working environment. This is attributed to interspecies differences and that the latency period for carcinogenicity is longer than acute toxicity studies as demonstrated in the experiments.
5. Conclusion
The systematic review analysed the association between exposure to TiO2 nanoparticle and lung cancer among TiO2 processing workers. The literature search utilized the QUT Library electronic data base which identified several TiO2 nanoparticle toxicity studies. From these studies only ten articles were extracted for analysis. Two articles out of the ten papers were cohort designed studies while the other eight were experimental designed studies. The two cohort studies revealed no association between respiratory related ailments and TiO2 exposure therefore having answered the review question of this paper. Although, the eight experimental studies demonstrated varying conclusions they applied common methodologies such as in using biological modelling, and did identify routes of TiO2 nanoparticle exposure as the agent for toxicity. Moreover, a number of the experimental studies have demonstrated cell toxicity in the form of cell necrosis, apoptotic manifestations and oxidative stress effects which were dose response dependent. These results reflect the likely exposure scenario to exposed workers in TiO2 processing plants. Although less is known about the health risks associated with TiO2 nanoparticle there is adequate evidence in animal and human cell studies to be concerned with which require appropriate Authorities to be proactive in developing risk mitigation controls against unnecessary exposure to TiO2 nanoparticle.