Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

An Analysis of Biopsies for Suspected Skin Cancer at a Tertiary Care Dermatology Clinic in the Western Cape Province of South Africa

An Analysis of Biopsies for Suspected Skin Cancer at a Tertiary Care Dermatology Clinic in the... Hindawi Journal of Skin Cancer Volume 2020, Article ID 9061532, 7 pages https://doi.org/10.1155/2020/9061532 Research Article An Analysis of Biopsies for Suspected Skin Cancer at a Tertiary Care Dermatology Clinic in the Western Cape Province of South Africa 1 2 1 3 1 Johann de Wet , Minette Steyn, Henry F. Jordaan, Rhodine Smith, Saskya Claasens, and Willem I. Visser Division of Dermatology, Department of Medicine, Tygerberg Academic Hospital, Stellenbosch University, Cape Town, South Africa Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa Division of Epidemiology and Biostatistics, Department of Global Health, Stellenbosch University, Cape Town, South Africa Correspondence should be addressed to Johann de Wet; dewetjohann@yahoo.com Received 18 September 2019; Revised 11 October 2019; Accepted 24 October 2019; Published 27 January 2020 Academic Editor: Arash Kimyai-Asadi Copyright © 2020 Johann de Wet et al. is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background. Skin cancer is a growing health concern worldwide. It is the most common malignancy in South Africa and places a large burden on the public healthcare sector. ere is a paucity of published scientific data on skin cancer in South Africa. Objectives. To report the findings of biopsies performed in patients with suspected skin cancer attending the Tygerberg Academic Hospital (TAH) Dermatology outpatient department (OPD) in the Western Cape Province of South Africa. Methodology: A retrospective chart review identified all patients who underwent a biopsy for a suspected skin cancer diagnosis between September 2015 and August 2016 at the TAH dermatology OPD. Results. A total number of 696 biopsies from 390 participants were identified, of which 460 were histologically confirmed as malignant lesions. e proportion of clinically suspected skin cancers that were histologically confirmed as cancer was 68%. e most commonly occurring malignancies were basal cell carcinoma (BCC) (54.8%), squamous cell carcinoma (SCC) (18.9%), squamous cell carcinoma in-situ (SCCI) (8.0%), Kaposi’s sarcoma (KS) (6.7%), malignant melanoma (MM) (6.1%), and keratoacanthoma (KA) (4.6%). e number needed to treat (NTT) for all cancers diagnosed and for MM was 1.5 and 4 respectively. BCC (89.3%) and KS (67.7%) was the most common skin cancer in the white and black population respectively. e ratio of BCC to SCC was 2.03. Conclusion. is study provides valuable scientific data on the accuracy of skin cancer diagnosis, distribution and patient demographics in the Western Cape Province of South Africa, on which further research can be based. e study highlights the burden of skin cancer on this specific population group and calls for standardised reporting methods and increased surveillance of skin cancers. level of ultraviolet radiation (UVR) as well as the increased 1. Introduction sun exposure due to occupational and recreational activities Skin cancer is the most common malignancy worldwide and [8, 9]. While incidence rates for NMSC in South Africa remain is typically divided into MM and non-melanoma skin cancer high, rates are oen grossly underreported due to incomplete (NMSC) [1]. e term NMSC encompasses BCC and SCC but case reporting to the National Cancer Registry (NCR). It is also includes rarer types of skin cancer such as Merkel cell estimated that 25.4% of all cancers diagnosed in men in South carcinoma and KS [2]. e worldwide incidence of NMSC and Africa in 2014 were BCC while 10.9% were SCC [10]. In MM has been steadily increasing [3–7]. females, the numbers were 18.6% and 7.1% for BCCs and Most of the studies on NMSC have focussed on white pop- SCCs respectively [10]. is shows an increase when compared ulations in Europe, the United States (US), and Australia with to numbers from 2009 [11]. Although data on MM incidence limited data for other skin types in regions such as Africa [1]. on the African continent remain scarce, a recent study showed e South African population is at a particularly high risk the overall incidence of MM for South Africa to be 2.7 per of skin cancer due to the country’s geographical position and 100 000, with the incidence in the white population being 2 Journal of Skin Cancer T 1: Diagnosis of all biopsies. significantly higher at 23.2 per 100 000 [12]. MM in darker skin types oen presents as acral melanoma (AM), a rare BCC 252 (36.21) distinct variant of MM that arises from the palms, soles, and SCC 87 (12.5) nail apparatus and is associated with a poorer prognosis SCCI 37 (5.32) [13–15]. KS 31 (4.45) While skin type and UVR play a large role in determining Melanoma 28 (4.02) the risk of developing skin cancer, there are also many other Keratoacanthoma 21 (3.02) factors at play, including immunosuppression. e prevalence Histological diagnoses Other malignant 4 (0.57) of Human Immunodeficiency Virus (HIV) in South Africa (푁 = 696 )  (%) Solar keratosis 6 (0.86) was estimated at 12.7% in 2016 making the total number of Melanocytic naevi 21 (3.02) people living with HIV (PLHIV) approximately 7.03 million Seborrhoeic keratosis 29 (4.17) [16, 17]. Despite the rollout of the antiretroviral therapy (ART) programme in South Africa in 2004 the risk of KS amongst Dermatofibroma 5 (0.72) PLHIV remains elevated even in the ART era [18]. Omland Other benign 129 (18.53) et al. also observed a 2-fold increased risk of BCC and a 5-fold Insufficient biopsy/no 20 (2.87) increased risk of SCC in PLHIV compared with the back- histological dx ground population [19]. e increasing incidence of skin cancers places major approximately 1.4 million surnames of known ethnicity. is financial strain on South Africa’s already overburdened public method was constructed for the NCR as part of a Statistical healthcare system. A recent study estimated the annual cost Analysis Soware program by the Data Management and of skin cancer treatment in the country to be ZAR 92.4 million Statistical Analysis Unit of the University of the Witwatersrand ($15.7 million) with a further ZAR 45.1 million ($7.7 million) [21]. e NNT for all cancers refers to the number of biopsies spent on the workup of suspicious lesions that were ultimately performed to make the diagnoses of one skin cancer (all diagnosed as benign [20]. biopsies performed for skin cancer divided by the number of e above emphasises the growing need for accurate data skin cancers histologically confirmed). We calculated the NNT capturing of skin cancer in South Africa to promote and assist for MM, as the total number of pigmented lesions biopsied research, as well as to increase awareness regarding prevention. (where the clinician indicated melanoma as a provisional is study aimed to address the gap in the availability of data diagnosis) divided by the number of histologically confirmed and can pave the way for other research to be done on skin MM. Stata version 14 was used for data analysis. e analysis cancers at other similar clinics in South Africa. was of a descriptive nature. A biostatistician was consulted to assist with data analysis. e study was performed in accordance with ethical prin- 2. Objectives ciples in the Declaration of Helsinki and Good Clinical e primary objective was to assess the number of biopsies Practice. It was approved by the Health Research Ethics performed and determine the frequency and spectrum of his- Committee of the Faculty of Medicine and Health Sciences, tologically confirmed skin cancer. e secondary objectives University of Stellenbosch (HREC/REF: U16/10/028). All data were to (i) describe the demographics of patients diagnosed collected was held under the provisions of the 2013 Protection with skin cancer, (ii) duration and location of skin lesions of Personal Information Act (SA) and stored in secure manual biopsied and (iii) determine the accuracy of skin cancer diag- and electronic files. nosis by calculating the number needed to treat (NNT). 4. Results 3. Materials and Methods A total number of 1 444 biopsies were performed over a one- A retrospective study of a descriptive nature was conducted. year period (September 2015–August 2016) at the dermatol- It consisted of patients who received a biopsy specifically to ogy OPD at TAH, of which 696 biopsies, from 390 unique diagnose skin cancer (MM and NMSC) at the dermatology participants, met the inclusion criteria. OPD at TAH between 1st September 2015 and 31st August Of the 696 biopsies performed on clinically suspected skin 2016. e exclusion criteria entailed the following: (i) patients cancers, 460 (66.1%) were histologically confirmed skin can- under 18 years of age (ii), biopsies done for primary cutaneous cers, 216 (31%) were reported as benign lesions and in 20 cases T-cell lymphoma. Patients were identified using the biopsy (2.9%) a histological diagnosis could not be made (Table 1). registry of the dermatology OPD, TAH. ese biopsies were e percentage of lesions tested for skin cancer and confirmed performed by registrars and consultants in the field of histologically was 68%, yielding an NNT of 1.5 for all skin dermatology. Demographic, clinical, and histological data cancers. e NNT for melanoma was 4. e baseline demo- were collected from pathology reports from the National graphics of participants where skin cancer was confirmed Health Laboratory Service at TAH. A hot-deck imputation histologically is summarised in Table 2. method was used for subjects where ethnicity was not In 86.3% of the histologically confirmed skin cancers the indicated. Subjects were assigned to a certain ethnic group by exact histological diagnosis was included in the provisional comparing their surnames with a reference database of differential diagnosis by the clinician. Journal of Skin Cancer 3 T 2:  Demographics of patients with biopsy confirmed skin T 3: Location of all biopsies performed. cancer. Face 316 (48.47) Age (mean, 95% CI) 65,68 (63, 91-67, 46) Leg 93 (14.26) Male 142 (53.38%) Arm 53 (8.13) Sex (푁 = 390 )  (%) Female 124 (46.62%) Back 53 (8.13) Unknown 7 (2.63) Chest 30 (4.60) Mixed ancestry 34 (12.78) Hand 28 (4.29) Location of lesion (푁 = 652 )  (%) White 201 (75.56) Neck 26 (4.00) Race (푁 = 390 )  (%) Black 23 (8.65) Shoulders 21 (3.22) Asian 1 (0.38) Foot 12 (1.84) Indian 0 Abdomen 10 (1.53) Other 8 (1.23) Buttocks 2 (0.31) e commonest method of tumor sampling was punch biopsy (50%) followed by curette (10.1%), excision biopsy (9.1%), and then shave biopsy (1%) while the method used reports BCC as the most common cutaneous malignancy was unspecified in 29.9% of biopsies done. [1, 2, 23]. However, studies conducted in sub-Saharan Africa e most frequently biopsied body site was the face fol- concluded that SCC was the most common malignancy fol- lowed by the extremities and trunk (Table 3). lowed by KS. e largest review reported SCC and KS to con- e most commonly occurring malignancies were BCC stitute 44% and 25% respectively of the cutaneous malignancies (54.8%), SCC (18.9%), SCCI (8%), KS (6.7%), MM (6.1%), KA and BCC’s only 7% [24, 25]. Interestingly a recent study done (4.6%), and other malignant (0.9%). e ratio of BCC to SCC in the Northern Cape Province of South Africa also found that was 2.03. 45.4% of skin cancers diagnosed were SCC while only 27.8% Most lesions confirmed as malignancies were present were BCC [26]. ese differences in skin cancer frequency can between 1 and 6 months at the time of biopsy as reported by most likely be attributed to the difference in population dis- the patient. tribution in the Western Cape when compared to that of other e most frequent skin cancer per ethnic group is sum- sub-Saharan African countries and other parts of South Africa. marised in Table 4 and the age and gender distribution per e largest population group residing in the Western Cape skin cancer in Table 5. identifies as mixed ancestry with the population distribution being as follows: mixed ancestry (47.5%), black (35.7%), white (16.0%), and Indian/Asian (0.8%) [27]. In contrast the largest 5. Discussion population group in other parts of South Africa and sub-Sa- During the study period 48.2% of all biopsies performed at haran Africa will identify as black. Skin cancer prevalence the dermatology OPD at TAH were aimed at confirming or differs in populations of different skin types. BCC is the most excluding a diagnosis of skin cancer. is finding highlights common skin cancer in white, Hispanic, and Asian (Japanese the burden that skin cancer places on a tertiary dermatology and Chinese) populations, as is reflected in this current study, clinic in the public health care system in the Western Cape while SCC is the most prevalent skin cancer amongst black Province of South Africa. e number may even be an under- and Indian/Asian populations [28, 29]. estimation of the prevalence of skin cancer in the study pop- e ratio of BCC’s to SCC’s in our cohort was 2.03. Recent ulation as NMSC are occasionally diagnosed clinically and studies point to an increasing SCC incidence relative to BCC, treated without a biopsy [1]. moving the historical 4 : 1 ratio to 2.5 : 1 or even closer. Studies Of all skin lesions biopsied 45.5% of lesions occurred on attributes this to a relative SCC increase in the elderly popu- the face reflecting the propensity for skin cancer to affect sun- lation caused by chronic exposure to UVR [30]. In our study exposed sites. A recent British Association of Dermatologists a larger black population may be attributing to this National Audit on NMSC Excision also reported excisions phenomenon. from the head and neck to account for the majority (56.7%) MM accounted for 6% of all skin cancers diagnosed. is of cases [22]. is slightly higher than previous reports from the US and With regards to the demographics of subjects with a con- Australia that indicate MM to constitute approximately 4% firmed diagnosis of skin cancer it is not surprising that the and 2% of all skin cancers respectively [31, 32]. Previous stud- majority of patients were identified as white patients, as this ies in the Western Cape of South Africa reported an increased group will be most at risk of developing skin cancer [2, 12] incidence of MM in Cape Town specifically when compared (Table 2). e mean age for patients with confirmed skin can- to other parts of South Africa [33]. e number of MM cases cer was 65.68 and in keeping with previous reports on age for in our study may even be underrepresented since a large per- MM and NMSC that showed people over the age of 60 to have centage of MM are diagnosed in the private sector and possibly a higher incidence of skin cancer [2, 12]. at primary care clinics and secondary hospitals without being In our study cohort the most common malignancies in referred to a tertiary hospital for biopsy. Tod et al. found that descending order were BCC, SCC, SCCI, MM, KS, and KA. 75% of all MM cases in South Africa are diagnosed in the is is in keeping with international data on skin cancer, which private sector [12]. ere was only one case of MM diagnosed 4 Journal of Skin Cancer T 4: Skin cancer per ethnic group. Type of lesion White (, %) Mixed ancestry (, %) Black (, %) Asian (, %) Unknown (, %) Total 225 (89.29) 20 (7.94) 4 (1.59) 1 (0.40) 2 (0.79) 252 BCC 74 (85.06) 6 (6.90) 6 (6.90) 0 1 (1.15) 87 SCC 31 (83.78) 5 (13.51) 1 (2.70) 0 0 37 SCCI 3 (9.68) 3 (9.68) 21 (67.74) 0 4 (12.90) 31 KS 24 (85.71) 3 (10.71) 1 (3.57) 0 0 28 MM 18 (85.71) 2 (9.52) 0 0 1 (4.76) 21 KA 2 (50) 2 (50) 0 0 0 4 Malignant other Total 377 41 33 1 8 460 T 5: Age and gender distribution for all skin cancers. Skin cancer  (%) Age (mean, 95% CI) Male (, %) Female (, %) 252 (54.8) 68.53 (67.06–69.99) 154 (61.11) 98 (38.89) BCC 87 (18.9) 66.9 (63.76–70.04) 54 (62.07) 33 (37.93) SCC 37 (8) 66.9 (63.63–70.34) 21 (56.76) 16 (43.24) SCCI 31 (6.7) 41.29 (36.55–46.02) 15 (48.39) 16 (51.61) KS 28 (6.1) 68.13 (64.41–71.85) 13 (46.43) 15 (53.57) MM 21 (4.6) 66.46 (61.36–71.57) 11 (52.38) 10 (47.62) KA Malignant other 4 (0.9) 72.32 (61.57–83.06) 2 (50) 2 (50) in a patient from the black population. is lesion was located some authors consider KA as a precursor of SCC, a well-dif- on the foot, which is in keeping with the fact that AM is the ferentiated SCC or an abortive malignancy with invasive most common subtype in the black population [34]. A recent potential [35]. e true incidence of KA is probably underes- study done in the Western Cape Province reported that up to timated because of misdiagnosis as a SCC, underreporting or 22% of all MM diagnosed were AM and accounted for 80% of spontaneous regression [36]. e SCC/KA ratio in our cohort all MM diagnosed in black patients [15]. was 5.9. When compared to international published literature KS accounted for 7% of all skin cancers diagnosed and was this ratio varies in studies between 2.5 : 1 and 139 : 1, most the most frequent skin cancer in the black population. is likely due to the variation in approach by pathologists [36]. finding supports data from the Northern Cape Province of e proportion of clinically suspected skin cancers that South Africa that showed KS to account for 6.5% of all skin were histologically confirmed was 68%. is number com- cancers and occurring more commonly in the black popula- pares favourably to the Skin Cancer Audit & Research Database tion [26]. When compared to data from sub-Saharan African, (SCARD) of Australia which showed the percentage of new Nthumba et al. reported KS to represent 25% of skin cancers lesions tested which were malignant, during the same time diagnosed in Kenya in 2008 [24]. Although HIV status for period, to be 61%. SCARD is a surgical log designed for doc- subjects was not reported in the current study HIV infection tors treating skin malignancies in Australia with 122 562 remains the biggest risk factor for developing KS in South lesions biopsied during the same time period as the current Africa [18]. study (September 2015–August 2016) [37]. It is estimated that 18.9% of the South African population A recent publication that reported on skin biopsies and between the ages of 15–49 are HIV positive [16, 17]. e skin cancer treatment procedures in the US, showed that in higher rate of infection in this age group is reflected by the 2015 only about 50% of biopsies resulted in a skin cancer diag- mean age of presentation of KS in our study being 41.3 (Table nosis. e authors concluded that the threshold for biopsy 5). A study investigating the burden of cancers associated with may be decreasing, with more biopsies yielding negative HIV in the South African public health sector between 2004 results being performed [38]. and 2014 also demonstrated that cancer proportions were e NNT for all skin cancers in our study cohort was 1.5 highest between the ages of 25 and 49 with the greater pro- and is comparable to a study evaluating the US dermatologists portion of cancers observed in the black population. e Odds at discriminating skin cancers that showed an NNT for all skin Ratio for PLHIV to develop KS in SA was reported as 134 [18]. cancers of 2.22 [39]. PLHIV are also at higher risk for developing NMSC with a e NNT for melanoma was 4, referring to the number of 5-fold increased risk of SCC [18, 19]. is may also explain pigmented lesions needed to be biopsied to diagnose one mel- the lower BCC to SCC ratio as mentioned. anoma. If this is compared to SCARD data from the same time KA was considered a separate entity from SCC in the study period, lesions tested to find one melanoma was 5.3 [37]. cohort and comprised of 4% of skin cancers. e true nature Internationally published literature reported NNT numbers and its relationship to SCC continues to be controversial as for melanoma to vary from 4 to 40, with lower NNT’s generally Journal of Skin Cancer 5 being documented by dermatologists [40, 41]. Two similar the accuracy of databases and to promote and assist further studies in the United Kingdom (UK) analysing clinical diag- research. It emphasises the need to further describe risk factors nosis of melanoma by dermatologists reported NNT’s between for skin cancer, to promote prevention and to improve diag- 2.74 and 6.3 [42, 43]. e NNT in our study though was cal- nostic and management strategies. culated by including all pigmented lesion that were biopsied to rule out melanoma and not just melanocytic nevi, as was Data Availability the case in the UK studies. At the time of our study, dermoscopy was not routinely e data used to support the findings of this study are available utilized in the diagnosis of skin cancer at the TAH dermatol- from the corresponding author upon request. ogy OPD. Studies have shown that dermoscopy improves the diagnostic accuracy of both MM and NMSC leading to a decrease in NNT [44]. One can expect that the NNT for skin Conflicts of Interest cancers has improved at the TAH dermatology OPD since, as dermoscopy is now used routinely on every skin cancer e authors declare that they have no conflicts of interest. patient, although this remains to be tested. Punch biopsy was the mostly frequently utilized modality for skin cancer biopsies and accounted for 50% of all biopsies Authors’ Contributions performed. is contrasts with data from Australia that Johann de Wet and Minette Steyn contributed equally as first showed shave biopsies to make up 44.1% of biopsies compared authors. to 23.4% punch biopsies during the same time frame [37]. Farberg et al. who investigated the practice patterns of the US dermatologists for biopsy of MM also reported shave biopsy Acknowledgments (35%) to be the most commonly used method [45]. Shave biopsy offers the advantage of shorter procedure time, We acknowledge the contribution of the Division of decreased cost, and minimal bleeding [46]. In the past shave Anatomical Pathology, Faculty of Medicine and Health biopsy of cutaneous lesions has been discouraged out of fear Sciences, Stellenbosch University and National Health for compromising accurate diagnosis and microstaging of Laboratory Service at Tygerberg Hospital, Cape Town, South melanoma. Recent studies though have shown that when done Africa. correctly by a trained professional, shave biopsies are reliable and accurate in the majority of cases [47]. Training practices at the TAH may need to be reviewed to encourage shave biop- References sies for diagnosis of skin cancer. [1] A. Lomas, J. Leonardi-Bee, and F. Bath-Hextall, “A systematic review of worldwide incidence of nonmelanoma skin cancer,” 6. Study Limitations British Journal of Dermatology, vol. 166, no. 5, pp. 1069–1080, is was a retrospective study that relied on the completeness [2] V. Madan, J. T. Lear, and R. M. Szeimies, “Non-melanoma skin of pathology reports. Limiting factors were details such as cancer,” Lancet, vol. 375, no. 9715, pp. 673–685, 2010. duration, location, and ethnicity oen being omitted. Despite [3] J. F. Aitken, D. R. Youlden, P. D. Baade, H. P. Soyer, A. C. providing imputated ethnicity according to surname, there is Green, and B. M. Smithers, “Generational shi in melanoma no record of Fitzpatrick skin type, which presents difficulties incidence and mortality in Queensland, Australia, 1995–2014,” in deducing prevalence of certain malignancies in different International Journal of Cancer, vol. 142, no. 8, pp. 1528–1535, skin types. It should also be kept in mind that data collection was limited to one tertiary hospital and that the study took [4] American Cancer Society, “Cancer facts and figures,” 2019, place on a Western Cape population. e distribution of skin https://www.cancer.org/research/cancer-facts-statistics/all- types in the Western Cape is very different from the rest of cancer-facts-figures/cancer-facts-figures-2019.html (accessed South Africa and there is no previous data from the Western 15 July 2019). Cape from which comparisons could be drawn. In the light of [5] American Cancer Society, “Cancer Facts and Figures,” 2008, the high HIV prevalence in South Africa, the fact that HIV https://www.cancer.org/content/dam/cancer-org/research/ status of patients was not recorded leads to study limitation. c a n cer-fac ts-a n d-s t at i s t ics/a nnu a l-c a n cer-fac ts-a n d- figures/2008/cancer-facts-and-figures-2008.pdf (accessed 15 July 2019). 7. Conclusion [6] S. V. Mohan and A. L. Chang, “Advanced basal cell carcinoma: epidemiology and therapeutic innovations,” Current e study highlights the burden that skin cancer places on the Dermatology Reports, vol. 3, no. 1, pp. 40–45, 2014. health care system and provides valuable scientific data on the [7] H. W. Rogers, M. A. Weinstock, S. R. Feldman, and B. M. accuracy of skin cancer diagnosis, distribution, and patient Coldiron, “Incidence estimate of nonmelanoma skin cancer demographics in the Western Cape Province of South Africa. (keratinocyte carcinomas) in the US population, 2012,” JAMA e study calls for standardised reporting methods to increase Dermatology, vol. 151, no. 10, pp. 1081–1086, 2015. 6 Journal of Skin Cancer [8] C. W right, M. Norval, B. Summers, L. Davids, G. Coetzee, and [24] P. Nthumba, P. Cavadas, and L. Landin, “Primary cutaneous M. Oriowo, “Solar ultraviolet radiation exposure and human malignancies in Sub-Saharan Africa,” Annals of Plastic Surgery, health in South Africa: finding a balance,” South African Medical vol. 66, no. 3, pp. 313–320, 2011. Journal, vol. 102, no. 8, p. 665, 2012. [25] M. Asuquo, O. Ngim, G. Ugare, J. Omotoso, and G. Ebughe, “Major dermatologic malignancies encountered in a teaching [9] J. Russak and D. Rigel, “Risk factors for the development of hospital surgical department in South Nigeria,” American primary cutaneous melanoma,” Dermatologic Clinics, vol. 30, Journal of Clinical Dermatology, vol. 9, no. 6, pp. 383–387, 2008. no. 3, pp. 363–368, 2012. [26] K. York, N. Dlova, C. Wright et al., “Primary cutaneous [10] N ational Cancer Registry data, 2014, http://www.nicd.ac.za/wp- malignancies in the Northern Cape Province of South Africa: content/uploads/2017/03/2014-NCR-tables-1.pdf (accessed 15 a retrospective histopathological review,” South African Medical July 2019). Journal, vol. 107, no. 1, p. 83, 2016. [11] National Cancer Registry data, 2009, http://www.nioh.ac.za/ [27] Statistics South Africa, “Provincial profile: Western Cape wp-content/uploads/2018/03/NCR-2009-results.pdf (accessed Community Survey 2016,” Pretoria: Statistics South Africap. 16, 15 July 2019). 2018, http://cs2016.statssa.gov.za/wp-content/uploads/2016/07/ [12] B. M. Tod, P. E. Kellett, E. Singh, W. I. Visser, C. J. Lombard, and NT-30-06-2016-RELEASE-for-CS-2016- _Statistical-releas_1- C. Y. Wright, “e incidence of melanoma in South Africa: An July-2016.pdf (accessed 15 July 2019). exploratory analysis of National Cancer Registry data from 2005 [28] H. Gloster and K. Neal, “Skin cancer in skin of color,” Journal to 2013 with a specific focus on melanoma in black Africans,” of the American Academy of Dermatology, vol. 55, no. 5, South African Medical Journal, vol. 109, no. 4, pp. 246–253, pp. 741–760, 2006. [29] M. Gohara, “Skin cancer: an African perspective,” British [13] M. Norval and C. Y. Wright, “e epidemiology of cutaneous Journal of Dermatology, vol. 173, pp. 17–21, 2015. melanoma in the white and black African population groups in South Africa,” in Cutaneous Melanoma: Etiology and [30] M. C. Cameron, E. Lee, B. P. Hibler et al., “Basal cell carcinoma: erapy, W. H. Ward and J. M. Farma, Eds., pp. 23–38, Codon Epidemiology; pathophysiology; clinical and histological Publications, Brisbane, 2017. subtypes; and disease associations,” Journal of the American Academy of Dermatology, vol. 80, no. 2, pp. 303–317, 2019. [14] F . Durbec, L. Martin, C. Derancourt, and F. Grange, “Melanoma of the hand and foot: epidemiological, prognostic and genetic [31] R. Siegel, D. Naishadham, and A. Jemal, “Cancer statistics, features: a systematic review,” British Journal of Dermatology, 2013,” CA: A Cancer Journal for Clinicians, vol. 63, no. 1, vol. 166, no. 4, pp. 727–739, 2012. pp. 11–30, 2013. [15] J . De Wet, B. Tod, W. I. Visser, H. F. Jordaan, and J. W. Schneider, [32] Australian Institute of Health and Welfare 2017, “Cancer in “Clinical and pathological features of acral melanoma in a Australia 2017,” AIHW, Canberra, https://www.aihw.gov.au/ South African population: a retrospective study,” South African reports/cancer/cancer-in-australia-2017/contents/table-of- Medical Journal, vol. 108, no. 9, pp. 777–781, 2018. contents Cancer series no. 101. Cat. no. CAN 100 (accessed 15 July 2019). [16] Statistics South Africa, “Mid-year population estimates,” Pretoriap. 22, 2018. http://www.statssa.gov.za/publications/ [33] S. Jessop, H. Stubbings, R. Sayed, J. Duncan-Smith, J. Schneider, P0302/P03022018.pdf (accessed 15 July 2019). and H. Jordaan, “Regional clinical registry data show increased incidence of cutaneous melanoma in Cape Town,” South African [17] O. Shisana, T. Rehle, L. Simbayi, K. Zuma, S. Jooste, and S. Medical Journal, vol. 98, no. 3, pp. 197–199, 2008. Zungu, “South African National HIV prevalence, incidence and behaviour survey, 2012,” Human Sciences Research Council, [34] J. Lodder, W. Simson, and P. Becker, “Malignant melanoma of Cape Town, 2014, http:// www.hsrc.ac.za/en/research-data/ the skin in black South Africans: a 15-year experience,” South view/6871 (accessed 15 July 2019). African Journal of Surgery, vol. 38, no. 3, pp. 76–79, 2010. [18] T. Dhokotera, J. Bohlius, A. Spoerri et al., “e burden of [35] I. C. Watanabe, R. F. Magalhães, A. M. de Moraes et al., cancers associated with HIV in the South African public health “Keratoacanthoma and keratoacanthoma-like squamous cell sector, 2004–2014: a record linkage study,” Infectious Agents and carcinoma: similar morphology but different pathogenesis,” Cancer, vol. 14, no. 1, pp. 2004–2014, 2019. Medicine (Baltimore), vol. 94, no. 23, p. e934, 2015. [19] S. H. Omland, M. G. Ahlström, J. Gersto et al., “Risk of skin [36] B . Kwiek and R. A. Schwartz, “Keratoacanthoma (KA): an update cancer in patients with HIV: a Danish nationwide cohort study,” and review,” Journal of the American Academy of Dermatology, Journal of the American Academy of Dermatology, vol. 79, no. 4, vol. 74, no. 6, pp. 1220–1233, 2016. pp. 689–695, 2018. [37] “Skin Cancer Audit & Research Database (SCARD),” https:// [20] L. Gordon, T. Elliott, C. Wright, N. Deghaye, and W. Visser, scard.skincanceraudit.com/ (accessed July 15 2019). “Modelling the healthcare costs of skin cancer in South Africa,” [38] D. M. Wang, F. C. Morgan, R. J. Besaw, and C. D. Schmults, BMC Health Services Research, vol. 16, no. 1, p. 113, 2016. “An ecological study of skin biopsies and skin cancer treatment [21] R. J. Little and D. B. Rubin, “e analysis of social science data procedures in the United States Medicare population, 2000 to with missing values,” in Modern Methods of Data Analysis, J. Fox 2015,” Journal of the American Academy of Dermatology, vol. 78, and J. S. Long, Eds., pp. 292–326, Sage Publications, London, no. 1, pp. 47–53, 2018. [39] R. L. Wilson, B. A. Yentzer, S. P. Isom, S. R. Feldman, and A. B. [22] D. J. Keith, A. P. Bray, A. Brain et al., “British Association of Fleischer Jr, “How good are US dermatologists at discriminating Dermatologists (BAD) national audit on non-melanoma skin skin cancers? A number-needed-to-treat analysis,” Journal of cancer excision 2016 in collaboration with the royal college of Dermatological Treatment, vol. 23, no. 1, pp. 65–69, 2012. pathologists,” Clinical and Experimental Dermatology, 2019. [40] H . M. Rolfe, “Accuracy in skin cancer diagnosis: a retrospective [23] U. Leiter, U. Keim, T. Eigentler et al., “Incidence, mortality, and study of an Australian public hospital dermatology department,” trends of nonmelanoma skin cancer in Germany,” Journal of Australasian Journal of Dermatology, vol. 53, no. 2, pp. 112–117, Investigative Dermatology, vol. 137, no. 9, pp. 1860–1867, 2017. 2012. Journal of Skin Cancer 7 [41] A. L. K. Chia, G. Simonova, B. Dutta, A. Lim, and S. Shumack, “Melanoma diagnosis: Australian dermatologists’ number needed to treat,” Australasian Journal of Dermatology, vol. 49, no. 1, pp. 12–15, 2008. [42] B. Esdaile, I. Mahmud, A. Palmer, and J. Bowling, “Diagnosing melanoma: how do we assess how good we are?” Clinical and Experimental Dermatology, vol. 39, no. 2, pp. 129–134, 2014. [43] S. Sidhu, O. Bodger, N. Williams, and D. L. Roberts, “e number of benign moles excised for each malignant melanoma: the number needed to treat,” Clinical and Experimental Dermatology, vol. 37, no. 1, pp. 6–9, 2012. [44] C. Sinz, P. Tschandl, C. Rosendahl et al., “Accuracy of dermatoscopy for the diagnosis of nonpigmented cancers of the skin,” Journal of the American Academy of Dermatology, vol. 77, no. 6, pp. 1100–1109, 2017. [45] A. S. Farberg and D. S. Rigel, “A comparison of current practice patterns of US dermatologists versus published guidelines for the biopsy, initial management, and follow up of patients with primary cutaneous melanoma,” Journal of the American Academy of Dermatology, vol. 75, no. 6, pp. 1193–1197.e1, 2016. [46] D . M. Elston, E. J. Stratman, and S. J. Miller, “Skin biopsy: biopsy issues in specific diseases,” Journal of the American Academy of Dermatology, vol. 74, no. 1, pp. 1–16, 2016. [47] J. S. Zager, S. N. Hochwald, S. S. Marzban et al., “Shave biopsy is a safe and accurate method for the initial evaluation of melanoma,” Journal of the American College of Surgeons, vol. 212, no. 4, pp. 454–460, 2011. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Skin Cancer Hindawi Publishing Corporation

An Analysis of Biopsies for Suspected Skin Cancer at a Tertiary Care Dermatology Clinic in the Western Cape Province of South Africa

Loading next page...
 
/lp/hindawi-publishing-corporation/an-analysis-of-biopsies-for-suspected-skin-cancer-at-a-tertiary-care-wd5cT5EuS8

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2020 Johann de Wet et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
2090-2905
eISSN
2090-2913
DOI
10.1155/2020/9061532
Publisher site
See Article on Publisher Site

Abstract

Hindawi Journal of Skin Cancer Volume 2020, Article ID 9061532, 7 pages https://doi.org/10.1155/2020/9061532 Research Article An Analysis of Biopsies for Suspected Skin Cancer at a Tertiary Care Dermatology Clinic in the Western Cape Province of South Africa 1 2 1 3 1 Johann de Wet , Minette Steyn, Henry F. Jordaan, Rhodine Smith, Saskya Claasens, and Willem I. Visser Division of Dermatology, Department of Medicine, Tygerberg Academic Hospital, Stellenbosch University, Cape Town, South Africa Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa Division of Epidemiology and Biostatistics, Department of Global Health, Stellenbosch University, Cape Town, South Africa Correspondence should be addressed to Johann de Wet; dewetjohann@yahoo.com Received 18 September 2019; Revised 11 October 2019; Accepted 24 October 2019; Published 27 January 2020 Academic Editor: Arash Kimyai-Asadi Copyright © 2020 Johann de Wet et al. is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background. Skin cancer is a growing health concern worldwide. It is the most common malignancy in South Africa and places a large burden on the public healthcare sector. ere is a paucity of published scientific data on skin cancer in South Africa. Objectives. To report the findings of biopsies performed in patients with suspected skin cancer attending the Tygerberg Academic Hospital (TAH) Dermatology outpatient department (OPD) in the Western Cape Province of South Africa. Methodology: A retrospective chart review identified all patients who underwent a biopsy for a suspected skin cancer diagnosis between September 2015 and August 2016 at the TAH dermatology OPD. Results. A total number of 696 biopsies from 390 participants were identified, of which 460 were histologically confirmed as malignant lesions. e proportion of clinically suspected skin cancers that were histologically confirmed as cancer was 68%. e most commonly occurring malignancies were basal cell carcinoma (BCC) (54.8%), squamous cell carcinoma (SCC) (18.9%), squamous cell carcinoma in-situ (SCCI) (8.0%), Kaposi’s sarcoma (KS) (6.7%), malignant melanoma (MM) (6.1%), and keratoacanthoma (KA) (4.6%). e number needed to treat (NTT) for all cancers diagnosed and for MM was 1.5 and 4 respectively. BCC (89.3%) and KS (67.7%) was the most common skin cancer in the white and black population respectively. e ratio of BCC to SCC was 2.03. Conclusion. is study provides valuable scientific data on the accuracy of skin cancer diagnosis, distribution and patient demographics in the Western Cape Province of South Africa, on which further research can be based. e study highlights the burden of skin cancer on this specific population group and calls for standardised reporting methods and increased surveillance of skin cancers. level of ultraviolet radiation (UVR) as well as the increased 1. Introduction sun exposure due to occupational and recreational activities Skin cancer is the most common malignancy worldwide and [8, 9]. While incidence rates for NMSC in South Africa remain is typically divided into MM and non-melanoma skin cancer high, rates are oen grossly underreported due to incomplete (NMSC) [1]. e term NMSC encompasses BCC and SCC but case reporting to the National Cancer Registry (NCR). It is also includes rarer types of skin cancer such as Merkel cell estimated that 25.4% of all cancers diagnosed in men in South carcinoma and KS [2]. e worldwide incidence of NMSC and Africa in 2014 were BCC while 10.9% were SCC [10]. In MM has been steadily increasing [3–7]. females, the numbers were 18.6% and 7.1% for BCCs and Most of the studies on NMSC have focussed on white pop- SCCs respectively [10]. is shows an increase when compared ulations in Europe, the United States (US), and Australia with to numbers from 2009 [11]. Although data on MM incidence limited data for other skin types in regions such as Africa [1]. on the African continent remain scarce, a recent study showed e South African population is at a particularly high risk the overall incidence of MM for South Africa to be 2.7 per of skin cancer due to the country’s geographical position and 100 000, with the incidence in the white population being 2 Journal of Skin Cancer T 1: Diagnosis of all biopsies. significantly higher at 23.2 per 100 000 [12]. MM in darker skin types oen presents as acral melanoma (AM), a rare BCC 252 (36.21) distinct variant of MM that arises from the palms, soles, and SCC 87 (12.5) nail apparatus and is associated with a poorer prognosis SCCI 37 (5.32) [13–15]. KS 31 (4.45) While skin type and UVR play a large role in determining Melanoma 28 (4.02) the risk of developing skin cancer, there are also many other Keratoacanthoma 21 (3.02) factors at play, including immunosuppression. e prevalence Histological diagnoses Other malignant 4 (0.57) of Human Immunodeficiency Virus (HIV) in South Africa (푁 = 696 )  (%) Solar keratosis 6 (0.86) was estimated at 12.7% in 2016 making the total number of Melanocytic naevi 21 (3.02) people living with HIV (PLHIV) approximately 7.03 million Seborrhoeic keratosis 29 (4.17) [16, 17]. Despite the rollout of the antiretroviral therapy (ART) programme in South Africa in 2004 the risk of KS amongst Dermatofibroma 5 (0.72) PLHIV remains elevated even in the ART era [18]. Omland Other benign 129 (18.53) et al. also observed a 2-fold increased risk of BCC and a 5-fold Insufficient biopsy/no 20 (2.87) increased risk of SCC in PLHIV compared with the back- histological dx ground population [19]. e increasing incidence of skin cancers places major approximately 1.4 million surnames of known ethnicity. is financial strain on South Africa’s already overburdened public method was constructed for the NCR as part of a Statistical healthcare system. A recent study estimated the annual cost Analysis Soware program by the Data Management and of skin cancer treatment in the country to be ZAR 92.4 million Statistical Analysis Unit of the University of the Witwatersrand ($15.7 million) with a further ZAR 45.1 million ($7.7 million) [21]. e NNT for all cancers refers to the number of biopsies spent on the workup of suspicious lesions that were ultimately performed to make the diagnoses of one skin cancer (all diagnosed as benign [20]. biopsies performed for skin cancer divided by the number of e above emphasises the growing need for accurate data skin cancers histologically confirmed). We calculated the NNT capturing of skin cancer in South Africa to promote and assist for MM, as the total number of pigmented lesions biopsied research, as well as to increase awareness regarding prevention. (where the clinician indicated melanoma as a provisional is study aimed to address the gap in the availability of data diagnosis) divided by the number of histologically confirmed and can pave the way for other research to be done on skin MM. Stata version 14 was used for data analysis. e analysis cancers at other similar clinics in South Africa. was of a descriptive nature. A biostatistician was consulted to assist with data analysis. e study was performed in accordance with ethical prin- 2. Objectives ciples in the Declaration of Helsinki and Good Clinical e primary objective was to assess the number of biopsies Practice. It was approved by the Health Research Ethics performed and determine the frequency and spectrum of his- Committee of the Faculty of Medicine and Health Sciences, tologically confirmed skin cancer. e secondary objectives University of Stellenbosch (HREC/REF: U16/10/028). All data were to (i) describe the demographics of patients diagnosed collected was held under the provisions of the 2013 Protection with skin cancer, (ii) duration and location of skin lesions of Personal Information Act (SA) and stored in secure manual biopsied and (iii) determine the accuracy of skin cancer diag- and electronic files. nosis by calculating the number needed to treat (NNT). 4. Results 3. Materials and Methods A total number of 1 444 biopsies were performed over a one- A retrospective study of a descriptive nature was conducted. year period (September 2015–August 2016) at the dermatol- It consisted of patients who received a biopsy specifically to ogy OPD at TAH, of which 696 biopsies, from 390 unique diagnose skin cancer (MM and NMSC) at the dermatology participants, met the inclusion criteria. OPD at TAH between 1st September 2015 and 31st August Of the 696 biopsies performed on clinically suspected skin 2016. e exclusion criteria entailed the following: (i) patients cancers, 460 (66.1%) were histologically confirmed skin can- under 18 years of age (ii), biopsies done for primary cutaneous cers, 216 (31%) were reported as benign lesions and in 20 cases T-cell lymphoma. Patients were identified using the biopsy (2.9%) a histological diagnosis could not be made (Table 1). registry of the dermatology OPD, TAH. ese biopsies were e percentage of lesions tested for skin cancer and confirmed performed by registrars and consultants in the field of histologically was 68%, yielding an NNT of 1.5 for all skin dermatology. Demographic, clinical, and histological data cancers. e NNT for melanoma was 4. e baseline demo- were collected from pathology reports from the National graphics of participants where skin cancer was confirmed Health Laboratory Service at TAH. A hot-deck imputation histologically is summarised in Table 2. method was used for subjects where ethnicity was not In 86.3% of the histologically confirmed skin cancers the indicated. Subjects were assigned to a certain ethnic group by exact histological diagnosis was included in the provisional comparing their surnames with a reference database of differential diagnosis by the clinician. Journal of Skin Cancer 3 T 2:  Demographics of patients with biopsy confirmed skin T 3: Location of all biopsies performed. cancer. Face 316 (48.47) Age (mean, 95% CI) 65,68 (63, 91-67, 46) Leg 93 (14.26) Male 142 (53.38%) Arm 53 (8.13) Sex (푁 = 390 )  (%) Female 124 (46.62%) Back 53 (8.13) Unknown 7 (2.63) Chest 30 (4.60) Mixed ancestry 34 (12.78) Hand 28 (4.29) Location of lesion (푁 = 652 )  (%) White 201 (75.56) Neck 26 (4.00) Race (푁 = 390 )  (%) Black 23 (8.65) Shoulders 21 (3.22) Asian 1 (0.38) Foot 12 (1.84) Indian 0 Abdomen 10 (1.53) Other 8 (1.23) Buttocks 2 (0.31) e commonest method of tumor sampling was punch biopsy (50%) followed by curette (10.1%), excision biopsy (9.1%), and then shave biopsy (1%) while the method used reports BCC as the most common cutaneous malignancy was unspecified in 29.9% of biopsies done. [1, 2, 23]. However, studies conducted in sub-Saharan Africa e most frequently biopsied body site was the face fol- concluded that SCC was the most common malignancy fol- lowed by the extremities and trunk (Table 3). lowed by KS. e largest review reported SCC and KS to con- e most commonly occurring malignancies were BCC stitute 44% and 25% respectively of the cutaneous malignancies (54.8%), SCC (18.9%), SCCI (8%), KS (6.7%), MM (6.1%), KA and BCC’s only 7% [24, 25]. Interestingly a recent study done (4.6%), and other malignant (0.9%). e ratio of BCC to SCC in the Northern Cape Province of South Africa also found that was 2.03. 45.4% of skin cancers diagnosed were SCC while only 27.8% Most lesions confirmed as malignancies were present were BCC [26]. ese differences in skin cancer frequency can between 1 and 6 months at the time of biopsy as reported by most likely be attributed to the difference in population dis- the patient. tribution in the Western Cape when compared to that of other e most frequent skin cancer per ethnic group is sum- sub-Saharan African countries and other parts of South Africa. marised in Table 4 and the age and gender distribution per e largest population group residing in the Western Cape skin cancer in Table 5. identifies as mixed ancestry with the population distribution being as follows: mixed ancestry (47.5%), black (35.7%), white (16.0%), and Indian/Asian (0.8%) [27]. In contrast the largest 5. Discussion population group in other parts of South Africa and sub-Sa- During the study period 48.2% of all biopsies performed at haran Africa will identify as black. Skin cancer prevalence the dermatology OPD at TAH were aimed at confirming or differs in populations of different skin types. BCC is the most excluding a diagnosis of skin cancer. is finding highlights common skin cancer in white, Hispanic, and Asian (Japanese the burden that skin cancer places on a tertiary dermatology and Chinese) populations, as is reflected in this current study, clinic in the public health care system in the Western Cape while SCC is the most prevalent skin cancer amongst black Province of South Africa. e number may even be an under- and Indian/Asian populations [28, 29]. estimation of the prevalence of skin cancer in the study pop- e ratio of BCC’s to SCC’s in our cohort was 2.03. Recent ulation as NMSC are occasionally diagnosed clinically and studies point to an increasing SCC incidence relative to BCC, treated without a biopsy [1]. moving the historical 4 : 1 ratio to 2.5 : 1 or even closer. Studies Of all skin lesions biopsied 45.5% of lesions occurred on attributes this to a relative SCC increase in the elderly popu- the face reflecting the propensity for skin cancer to affect sun- lation caused by chronic exposure to UVR [30]. In our study exposed sites. A recent British Association of Dermatologists a larger black population may be attributing to this National Audit on NMSC Excision also reported excisions phenomenon. from the head and neck to account for the majority (56.7%) MM accounted for 6% of all skin cancers diagnosed. is of cases [22]. is slightly higher than previous reports from the US and With regards to the demographics of subjects with a con- Australia that indicate MM to constitute approximately 4% firmed diagnosis of skin cancer it is not surprising that the and 2% of all skin cancers respectively [31, 32]. Previous stud- majority of patients were identified as white patients, as this ies in the Western Cape of South Africa reported an increased group will be most at risk of developing skin cancer [2, 12] incidence of MM in Cape Town specifically when compared (Table 2). e mean age for patients with confirmed skin can- to other parts of South Africa [33]. e number of MM cases cer was 65.68 and in keeping with previous reports on age for in our study may even be underrepresented since a large per- MM and NMSC that showed people over the age of 60 to have centage of MM are diagnosed in the private sector and possibly a higher incidence of skin cancer [2, 12]. at primary care clinics and secondary hospitals without being In our study cohort the most common malignancies in referred to a tertiary hospital for biopsy. Tod et al. found that descending order were BCC, SCC, SCCI, MM, KS, and KA. 75% of all MM cases in South Africa are diagnosed in the is is in keeping with international data on skin cancer, which private sector [12]. ere was only one case of MM diagnosed 4 Journal of Skin Cancer T 4: Skin cancer per ethnic group. Type of lesion White (, %) Mixed ancestry (, %) Black (, %) Asian (, %) Unknown (, %) Total 225 (89.29) 20 (7.94) 4 (1.59) 1 (0.40) 2 (0.79) 252 BCC 74 (85.06) 6 (6.90) 6 (6.90) 0 1 (1.15) 87 SCC 31 (83.78) 5 (13.51) 1 (2.70) 0 0 37 SCCI 3 (9.68) 3 (9.68) 21 (67.74) 0 4 (12.90) 31 KS 24 (85.71) 3 (10.71) 1 (3.57) 0 0 28 MM 18 (85.71) 2 (9.52) 0 0 1 (4.76) 21 KA 2 (50) 2 (50) 0 0 0 4 Malignant other Total 377 41 33 1 8 460 T 5: Age and gender distribution for all skin cancers. Skin cancer  (%) Age (mean, 95% CI) Male (, %) Female (, %) 252 (54.8) 68.53 (67.06–69.99) 154 (61.11) 98 (38.89) BCC 87 (18.9) 66.9 (63.76–70.04) 54 (62.07) 33 (37.93) SCC 37 (8) 66.9 (63.63–70.34) 21 (56.76) 16 (43.24) SCCI 31 (6.7) 41.29 (36.55–46.02) 15 (48.39) 16 (51.61) KS 28 (6.1) 68.13 (64.41–71.85) 13 (46.43) 15 (53.57) MM 21 (4.6) 66.46 (61.36–71.57) 11 (52.38) 10 (47.62) KA Malignant other 4 (0.9) 72.32 (61.57–83.06) 2 (50) 2 (50) in a patient from the black population. is lesion was located some authors consider KA as a precursor of SCC, a well-dif- on the foot, which is in keeping with the fact that AM is the ferentiated SCC or an abortive malignancy with invasive most common subtype in the black population [34]. A recent potential [35]. e true incidence of KA is probably underes- study done in the Western Cape Province reported that up to timated because of misdiagnosis as a SCC, underreporting or 22% of all MM diagnosed were AM and accounted for 80% of spontaneous regression [36]. e SCC/KA ratio in our cohort all MM diagnosed in black patients [15]. was 5.9. When compared to international published literature KS accounted for 7% of all skin cancers diagnosed and was this ratio varies in studies between 2.5 : 1 and 139 : 1, most the most frequent skin cancer in the black population. is likely due to the variation in approach by pathologists [36]. finding supports data from the Northern Cape Province of e proportion of clinically suspected skin cancers that South Africa that showed KS to account for 6.5% of all skin were histologically confirmed was 68%. is number com- cancers and occurring more commonly in the black popula- pares favourably to the Skin Cancer Audit & Research Database tion [26]. When compared to data from sub-Saharan African, (SCARD) of Australia which showed the percentage of new Nthumba et al. reported KS to represent 25% of skin cancers lesions tested which were malignant, during the same time diagnosed in Kenya in 2008 [24]. Although HIV status for period, to be 61%. SCARD is a surgical log designed for doc- subjects was not reported in the current study HIV infection tors treating skin malignancies in Australia with 122 562 remains the biggest risk factor for developing KS in South lesions biopsied during the same time period as the current Africa [18]. study (September 2015–August 2016) [37]. It is estimated that 18.9% of the South African population A recent publication that reported on skin biopsies and between the ages of 15–49 are HIV positive [16, 17]. e skin cancer treatment procedures in the US, showed that in higher rate of infection in this age group is reflected by the 2015 only about 50% of biopsies resulted in a skin cancer diag- mean age of presentation of KS in our study being 41.3 (Table nosis. e authors concluded that the threshold for biopsy 5). A study investigating the burden of cancers associated with may be decreasing, with more biopsies yielding negative HIV in the South African public health sector between 2004 results being performed [38]. and 2014 also demonstrated that cancer proportions were e NNT for all skin cancers in our study cohort was 1.5 highest between the ages of 25 and 49 with the greater pro- and is comparable to a study evaluating the US dermatologists portion of cancers observed in the black population. e Odds at discriminating skin cancers that showed an NNT for all skin Ratio for PLHIV to develop KS in SA was reported as 134 [18]. cancers of 2.22 [39]. PLHIV are also at higher risk for developing NMSC with a e NNT for melanoma was 4, referring to the number of 5-fold increased risk of SCC [18, 19]. is may also explain pigmented lesions needed to be biopsied to diagnose one mel- the lower BCC to SCC ratio as mentioned. anoma. If this is compared to SCARD data from the same time KA was considered a separate entity from SCC in the study period, lesions tested to find one melanoma was 5.3 [37]. cohort and comprised of 4% of skin cancers. e true nature Internationally published literature reported NNT numbers and its relationship to SCC continues to be controversial as for melanoma to vary from 4 to 40, with lower NNT’s generally Journal of Skin Cancer 5 being documented by dermatologists [40, 41]. Two similar the accuracy of databases and to promote and assist further studies in the United Kingdom (UK) analysing clinical diag- research. It emphasises the need to further describe risk factors nosis of melanoma by dermatologists reported NNT’s between for skin cancer, to promote prevention and to improve diag- 2.74 and 6.3 [42, 43]. e NNT in our study though was cal- nostic and management strategies. culated by including all pigmented lesion that were biopsied to rule out melanoma and not just melanocytic nevi, as was Data Availability the case in the UK studies. At the time of our study, dermoscopy was not routinely e data used to support the findings of this study are available utilized in the diagnosis of skin cancer at the TAH dermatol- from the corresponding author upon request. ogy OPD. Studies have shown that dermoscopy improves the diagnostic accuracy of both MM and NMSC leading to a decrease in NNT [44]. One can expect that the NNT for skin Conflicts of Interest cancers has improved at the TAH dermatology OPD since, as dermoscopy is now used routinely on every skin cancer e authors declare that they have no conflicts of interest. patient, although this remains to be tested. Punch biopsy was the mostly frequently utilized modality for skin cancer biopsies and accounted for 50% of all biopsies Authors’ Contributions performed. is contrasts with data from Australia that Johann de Wet and Minette Steyn contributed equally as first showed shave biopsies to make up 44.1% of biopsies compared authors. to 23.4% punch biopsies during the same time frame [37]. Farberg et al. who investigated the practice patterns of the US dermatologists for biopsy of MM also reported shave biopsy Acknowledgments (35%) to be the most commonly used method [45]. Shave biopsy offers the advantage of shorter procedure time, We acknowledge the contribution of the Division of decreased cost, and minimal bleeding [46]. In the past shave Anatomical Pathology, Faculty of Medicine and Health biopsy of cutaneous lesions has been discouraged out of fear Sciences, Stellenbosch University and National Health for compromising accurate diagnosis and microstaging of Laboratory Service at Tygerberg Hospital, Cape Town, South melanoma. Recent studies though have shown that when done Africa. correctly by a trained professional, shave biopsies are reliable and accurate in the majority of cases [47]. Training practices at the TAH may need to be reviewed to encourage shave biop- References sies for diagnosis of skin cancer. [1] A. Lomas, J. Leonardi-Bee, and F. Bath-Hextall, “A systematic review of worldwide incidence of nonmelanoma skin cancer,” 6. Study Limitations British Journal of Dermatology, vol. 166, no. 5, pp. 1069–1080, is was a retrospective study that relied on the completeness [2] V. Madan, J. T. Lear, and R. M. Szeimies, “Non-melanoma skin of pathology reports. Limiting factors were details such as cancer,” Lancet, vol. 375, no. 9715, pp. 673–685, 2010. duration, location, and ethnicity oen being omitted. Despite [3] J. F. Aitken, D. R. Youlden, P. D. Baade, H. P. Soyer, A. C. providing imputated ethnicity according to surname, there is Green, and B. M. Smithers, “Generational shi in melanoma no record of Fitzpatrick skin type, which presents difficulties incidence and mortality in Queensland, Australia, 1995–2014,” in deducing prevalence of certain malignancies in different International Journal of Cancer, vol. 142, no. 8, pp. 1528–1535, skin types. It should also be kept in mind that data collection was limited to one tertiary hospital and that the study took [4] American Cancer Society, “Cancer facts and figures,” 2019, place on a Western Cape population. e distribution of skin https://www.cancer.org/research/cancer-facts-statistics/all- types in the Western Cape is very different from the rest of cancer-facts-figures/cancer-facts-figures-2019.html (accessed South Africa and there is no previous data from the Western 15 July 2019). Cape from which comparisons could be drawn. In the light of [5] American Cancer Society, “Cancer Facts and Figures,” 2008, the high HIV prevalence in South Africa, the fact that HIV https://www.cancer.org/content/dam/cancer-org/research/ status of patients was not recorded leads to study limitation. c a n cer-fac ts-a n d-s t at i s t ics/a nnu a l-c a n cer-fac ts-a n d- figures/2008/cancer-facts-and-figures-2008.pdf (accessed 15 July 2019). 7. Conclusion [6] S. V. Mohan and A. L. Chang, “Advanced basal cell carcinoma: epidemiology and therapeutic innovations,” Current e study highlights the burden that skin cancer places on the Dermatology Reports, vol. 3, no. 1, pp. 40–45, 2014. health care system and provides valuable scientific data on the [7] H. W. Rogers, M. A. Weinstock, S. R. Feldman, and B. M. accuracy of skin cancer diagnosis, distribution, and patient Coldiron, “Incidence estimate of nonmelanoma skin cancer demographics in the Western Cape Province of South Africa. (keratinocyte carcinomas) in the US population, 2012,” JAMA e study calls for standardised reporting methods to increase Dermatology, vol. 151, no. 10, pp. 1081–1086, 2015. 6 Journal of Skin Cancer [8] C. W right, M. Norval, B. Summers, L. Davids, G. Coetzee, and [24] P. Nthumba, P. Cavadas, and L. Landin, “Primary cutaneous M. Oriowo, “Solar ultraviolet radiation exposure and human malignancies in Sub-Saharan Africa,” Annals of Plastic Surgery, health in South Africa: finding a balance,” South African Medical vol. 66, no. 3, pp. 313–320, 2011. Journal, vol. 102, no. 8, p. 665, 2012. [25] M. Asuquo, O. Ngim, G. Ugare, J. Omotoso, and G. Ebughe, “Major dermatologic malignancies encountered in a teaching [9] J. Russak and D. Rigel, “Risk factors for the development of hospital surgical department in South Nigeria,” American primary cutaneous melanoma,” Dermatologic Clinics, vol. 30, Journal of Clinical Dermatology, vol. 9, no. 6, pp. 383–387, 2008. no. 3, pp. 363–368, 2012. [26] K. York, N. Dlova, C. Wright et al., “Primary cutaneous [10] N ational Cancer Registry data, 2014, http://www.nicd.ac.za/wp- malignancies in the Northern Cape Province of South Africa: content/uploads/2017/03/2014-NCR-tables-1.pdf (accessed 15 a retrospective histopathological review,” South African Medical July 2019). Journal, vol. 107, no. 1, p. 83, 2016. [11] National Cancer Registry data, 2009, http://www.nioh.ac.za/ [27] Statistics South Africa, “Provincial profile: Western Cape wp-content/uploads/2018/03/NCR-2009-results.pdf (accessed Community Survey 2016,” Pretoria: Statistics South Africap. 16, 15 July 2019). 2018, http://cs2016.statssa.gov.za/wp-content/uploads/2016/07/ [12] B. M. Tod, P. E. Kellett, E. Singh, W. I. Visser, C. J. Lombard, and NT-30-06-2016-RELEASE-for-CS-2016- _Statistical-releas_1- C. Y. Wright, “e incidence of melanoma in South Africa: An July-2016.pdf (accessed 15 July 2019). exploratory analysis of National Cancer Registry data from 2005 [28] H. Gloster and K. Neal, “Skin cancer in skin of color,” Journal to 2013 with a specific focus on melanoma in black Africans,” of the American Academy of Dermatology, vol. 55, no. 5, South African Medical Journal, vol. 109, no. 4, pp. 246–253, pp. 741–760, 2006. [29] M. Gohara, “Skin cancer: an African perspective,” British [13] M. Norval and C. Y. Wright, “e epidemiology of cutaneous Journal of Dermatology, vol. 173, pp. 17–21, 2015. melanoma in the white and black African population groups in South Africa,” in Cutaneous Melanoma: Etiology and [30] M. C. Cameron, E. Lee, B. P. Hibler et al., “Basal cell carcinoma: erapy, W. H. Ward and J. M. Farma, Eds., pp. 23–38, Codon Epidemiology; pathophysiology; clinical and histological Publications, Brisbane, 2017. subtypes; and disease associations,” Journal of the American Academy of Dermatology, vol. 80, no. 2, pp. 303–317, 2019. [14] F . Durbec, L. Martin, C. Derancourt, and F. Grange, “Melanoma of the hand and foot: epidemiological, prognostic and genetic [31] R. Siegel, D. Naishadham, and A. Jemal, “Cancer statistics, features: a systematic review,” British Journal of Dermatology, 2013,” CA: A Cancer Journal for Clinicians, vol. 63, no. 1, vol. 166, no. 4, pp. 727–739, 2012. pp. 11–30, 2013. [15] J . De Wet, B. Tod, W. I. Visser, H. F. Jordaan, and J. W. Schneider, [32] Australian Institute of Health and Welfare 2017, “Cancer in “Clinical and pathological features of acral melanoma in a Australia 2017,” AIHW, Canberra, https://www.aihw.gov.au/ South African population: a retrospective study,” South African reports/cancer/cancer-in-australia-2017/contents/table-of- Medical Journal, vol. 108, no. 9, pp. 777–781, 2018. contents Cancer series no. 101. Cat. no. CAN 100 (accessed 15 July 2019). [16] Statistics South Africa, “Mid-year population estimates,” Pretoriap. 22, 2018. http://www.statssa.gov.za/publications/ [33] S. Jessop, H. Stubbings, R. Sayed, J. Duncan-Smith, J. Schneider, P0302/P03022018.pdf (accessed 15 July 2019). and H. Jordaan, “Regional clinical registry data show increased incidence of cutaneous melanoma in Cape Town,” South African [17] O. Shisana, T. Rehle, L. Simbayi, K. Zuma, S. Jooste, and S. Medical Journal, vol. 98, no. 3, pp. 197–199, 2008. Zungu, “South African National HIV prevalence, incidence and behaviour survey, 2012,” Human Sciences Research Council, [34] J. Lodder, W. Simson, and P. Becker, “Malignant melanoma of Cape Town, 2014, http:// www.hsrc.ac.za/en/research-data/ the skin in black South Africans: a 15-year experience,” South view/6871 (accessed 15 July 2019). African Journal of Surgery, vol. 38, no. 3, pp. 76–79, 2010. [18] T. Dhokotera, J. Bohlius, A. Spoerri et al., “e burden of [35] I. C. Watanabe, R. F. Magalhães, A. M. de Moraes et al., cancers associated with HIV in the South African public health “Keratoacanthoma and keratoacanthoma-like squamous cell sector, 2004–2014: a record linkage study,” Infectious Agents and carcinoma: similar morphology but different pathogenesis,” Cancer, vol. 14, no. 1, pp. 2004–2014, 2019. Medicine (Baltimore), vol. 94, no. 23, p. e934, 2015. [19] S. H. Omland, M. G. Ahlström, J. Gersto et al., “Risk of skin [36] B . Kwiek and R. A. Schwartz, “Keratoacanthoma (KA): an update cancer in patients with HIV: a Danish nationwide cohort study,” and review,” Journal of the American Academy of Dermatology, Journal of the American Academy of Dermatology, vol. 79, no. 4, vol. 74, no. 6, pp. 1220–1233, 2016. pp. 689–695, 2018. [37] “Skin Cancer Audit & Research Database (SCARD),” https:// [20] L. Gordon, T. Elliott, C. Wright, N. Deghaye, and W. Visser, scard.skincanceraudit.com/ (accessed July 15 2019). “Modelling the healthcare costs of skin cancer in South Africa,” [38] D. M. Wang, F. C. Morgan, R. J. Besaw, and C. D. Schmults, BMC Health Services Research, vol. 16, no. 1, p. 113, 2016. “An ecological study of skin biopsies and skin cancer treatment [21] R. J. Little and D. B. Rubin, “e analysis of social science data procedures in the United States Medicare population, 2000 to with missing values,” in Modern Methods of Data Analysis, J. Fox 2015,” Journal of the American Academy of Dermatology, vol. 78, and J. S. Long, Eds., pp. 292–326, Sage Publications, London, no. 1, pp. 47–53, 2018. [39] R. L. Wilson, B. A. Yentzer, S. P. Isom, S. R. Feldman, and A. B. [22] D. J. Keith, A. P. Bray, A. Brain et al., “British Association of Fleischer Jr, “How good are US dermatologists at discriminating Dermatologists (BAD) national audit on non-melanoma skin skin cancers? A number-needed-to-treat analysis,” Journal of cancer excision 2016 in collaboration with the royal college of Dermatological Treatment, vol. 23, no. 1, pp. 65–69, 2012. pathologists,” Clinical and Experimental Dermatology, 2019. [40] H . M. Rolfe, “Accuracy in skin cancer diagnosis: a retrospective [23] U. Leiter, U. Keim, T. Eigentler et al., “Incidence, mortality, and study of an Australian public hospital dermatology department,” trends of nonmelanoma skin cancer in Germany,” Journal of Australasian Journal of Dermatology, vol. 53, no. 2, pp. 112–117, Investigative Dermatology, vol. 137, no. 9, pp. 1860–1867, 2017. 2012. Journal of Skin Cancer 7 [41] A. L. K. Chia, G. Simonova, B. Dutta, A. Lim, and S. Shumack, “Melanoma diagnosis: Australian dermatologists’ number needed to treat,” Australasian Journal of Dermatology, vol. 49, no. 1, pp. 12–15, 2008. [42] B. Esdaile, I. Mahmud, A. Palmer, and J. Bowling, “Diagnosing melanoma: how do we assess how good we are?” Clinical and Experimental Dermatology, vol. 39, no. 2, pp. 129–134, 2014. [43] S. Sidhu, O. Bodger, N. Williams, and D. L. Roberts, “e number of benign moles excised for each malignant melanoma: the number needed to treat,” Clinical and Experimental Dermatology, vol. 37, no. 1, pp. 6–9, 2012. [44] C. Sinz, P. Tschandl, C. Rosendahl et al., “Accuracy of dermatoscopy for the diagnosis of nonpigmented cancers of the skin,” Journal of the American Academy of Dermatology, vol. 77, no. 6, pp. 1100–1109, 2017. [45] A. S. Farberg and D. S. Rigel, “A comparison of current practice patterns of US dermatologists versus published guidelines for the biopsy, initial management, and follow up of patients with primary cutaneous melanoma,” Journal of the American Academy of Dermatology, vol. 75, no. 6, pp. 1193–1197.e1, 2016. [46] D . M. Elston, E. J. Stratman, and S. J. Miller, “Skin biopsy: biopsy issues in specific diseases,” Journal of the American Academy of Dermatology, vol. 74, no. 1, pp. 1–16, 2016. [47] J. S. Zager, S. N. Hochwald, S. S. Marzban et al., “Shave biopsy is a safe and accurate method for the initial evaluation of melanoma,” Journal of the American College of Surgeons, vol. 212, no. 4, pp. 454–460, 2011.

Journal

Journal of Skin CancerHindawi Publishing Corporation

Published: Jan 27, 2020

References