Squamous Cell Carcinomas of the Head and Neck in Fanconi Anemia: Risk, Prevention, Therapy, and the Need for Guidelines

Abstract

Fanconi anemia (FA) is a rare recessive DNA repair disorder that is clinically characterized by congenital malformations, progressive bone marrow failure, and increased incidence of malignancies, especially acute myeloid leukemia and squamous cell carcinomas of the head and neck (HNSCCs) and the anogenital regions. On a cellular level, typical features of the disorder are a high degree of genomic instability and an increased sensitivity to bi-functionally alkylating agents. So far, germ-line defects in 15 different FA genes have been identified. Some of these FA genes are also established as tumor susceptibility genes for familiar cancers.

In recent years, the prevention and therapy of HNSCCs in FA patients has become more important as the percentage of patients surviving into adulthood is rising. HNSCCs appear in very young FA patients without common risk factors. Since cisplatin-based chemotherapy in combination with radiotherapy, essential parts of the standard treatment approach for sporadic HNSCCs, cannot be used in FA patients due to therapy-associated toxicities and mortalities even with reduced dosing, surgery is the most important treatment option for HNSCCs, in FA patients and requires an early and efficient detection of malignant lesions. So far, no uniform treatment protocol for the management of HNSCCs in FA patients exists. Therefore, we propose that the information on affected FA patients should be collected world-wide, practical therapeutic guidelines developed and national treatment centers established.

Introduction

Fanconi anemia (FA) is an autosomal or X-chromosomal recessive chromosome instability disorder due to germ-line mutations in at least 15 DNA repair genes: FANCA/B/C/D1/D2/E/F/G/I/J/L//M/N/O/P [16,31]. Clinically, FA is characterized by congenital malformations, progressive bone marrow failure, endocrine abnormalities and a high propensity of developing malignancies early in life [35]. Although certain subtypes of FA, predominantly with biallelic mutations in FANCD1/BRCA2 and FANCN/PALB2, develop a widerange of malignancies including AML, ALL nephroblastoma, medulloblastoma, and Non-Hodgkin-Lymphoma in the first decade of life [48,52,76], heterozygous germ-line mutations in the FA genes FANCD1/BRCA2, FANCJ/BRIP1, FANCN/PALB2 and FANCO/RAD51C are a cause of hereditary susceptibility for gynecological and other tumors [18,29,50]. However, the high incidence of MDS/AML and HNSCCs in the second and third decade of life [35] in FA patients with bi-allelic mutations in classical FA genes, where heterozygous germ-line mutation carriers do not have an increased cancer incidence [50], points to the general importance of the FA/BRCA pathway for the maintenance of genome stability [16,31].

In the last 10 years, the results of treating the hematological problems in FA by allogeneic stem cell transplantation have dramatically improved, with long-term survival rates approaching 70% for matched unrelated and 90% for matched related donors [10, 21,27,40, 67,75]. This success appears to predominantly be due to the inclusion of fludarabin in the low-intensity conditioning regimens [10,21,40,67], the early detection of pre-leukemic changes and consecutively early transplantation [70], improvements in the prophylaxis of graft-versus-host disease (GvHD) and improved measures during the peritransplantation time period [40]. Independent from transplantation, androgens as an alternative treatment option for the failing hematopoietic system can improve survival at least in some patients well beyond 20 years of age and without the need for performing stem cell transplantation [58].

Therefore, with more FA patients surviving long-term and due to the fact that FA is a hereditary cancer susceptibility disorder, the spectrum of clinical problems is shifting towards the need for more efficient prophylaxis, earlier detection and also better treatment of these malignancies later in the life of FA patients. While this has partially been achieved for hematological malignancies [70], comparable strategies have still to be implemented and standardized for non-hematological tumors. In this overview, we will focus on squamous cell carcinomas of the head and neck region (HNSCC) as the second most frequent tumor entity in FA patients, partly promoted by the conditioning and immunological problems of the stem cell transplantation procedure.

HNSCC in FA

Only 6% of all human malignant tumors worldwide are squamous cell carcinomas of the head and neck region (HNSCCs) [19,51]. More than 50% of the HNSCCs have already progressed to locally advanced disease by the time of diagnosis, thus requiring aggressive multimodality approaches to achieve at least 40–50% 5-year survival [19,51]. Although the recent addition of high-dose cisplatin to surgery and local irradiation was important for lowering the rate of both local regrowth and systemic dissemination in sporadic HNSCCs, this multidisciplinary therapy has a high rate of severe adverse effects as well as an increased incidence of late effects [5,12].

In FA however, HNSCC is the most frequently diagnosed solid tumor [34,35]. In 2003, the 20-year perspective of the International FA Registry (IFAR) reported an average survival of FA patients of only <25 years [35]. HNSCCS were diagnosed in 19(3%) of 754 FA patients and half of the patients had died because of their HNSCC [34,35]. The mean tumor manifestation age, 31 years, was very young compared to the median age of diagnosis of cancer of the oral cavity and pharynx in the U.S.A. population in 2004–2008,62 years (http://seer.cancer.gov/statfacts/html/oral-cav.html). Primary tumors predominantly appeared in the oral cavity, mainly the tongue or the gingiva. Noteworthy is that approximately one third of the malignancies in the head and neck region are found in the oropharynx, nasopharynx, hypopharynx and larynx and therefore are not readily detectable by examinations of the oral cavity [2,34]. In addition, female FA patients might have a higher rate of esophageal cancers [2,34]. In 2011, a mean tumor diagnosis age of 35.5 years was reported in 12 FA patients; only 4 of these patients were tumor-free for an average of 24 months after the diagnosis of HNSCC [7].

Importantly, in one study [2], the cancers in the upper aerodigestive tract in a relatively high percentage of cases, 9 out of 42 patients, were diagnosed prior to the diagnosis of FA. This was probably due to hypomorphic mutations [11,25] or reversions/mosaicism in the hematopoietic system [3]. As FA cells are hypersensitive to standard genotoxic treatment approaches including the mono- and bifunctional alkylators such as cisplatin, mitomycin C and cyclophosphamide as well as irradiation, undiagnosed FA patients will experience severe complications under standard HNSCC therapy and develop lethal treatment-associated toxicities [7,11,25,69].

Based on a comparison between a North-American Survey group of nontransplanted (n = 145) and the Saint Louis Hospital-derived group of transplanted FA patients (n=117), it was calculated that approximately 50% of the nontransplanted FA patients will develop HNSCC by 45 years of age [56]. In contrast, the cumulative incidence of HNSCC was estimated at approximately 100% for transplanted FA patients at the same age [56]. Major transplant-associated risk factors that influence the manifestation of HNSCC and also other epithelial tumors are the conditioning regimen and the occurrence of acute and chronic GvHD [55]. Of 13 FA patients with HNSCC who were transplanted at the Saint Louis Hospital between 1976 and 2007, all patients had received irradiation-based conditioning and all had developed extensive chronic GvHD [42]. In this study, patients developed HNSCCs on average 10 years after transplantation, predominantly in the oral cavity. At the last reported follow-up, only 2 patients were alive 9 and 23 months after HNSCC diagnosis despite the use of surgery, radiation and/or chemotherapy [42].

In summary, the cumulative incidence to develop HNSCC in FA patients increases at a greater-than-linear rate, approaching 4.4% per year by age 40 for nontransplanted FA patients and 10.1% per year >15 years after transplantation [56]. Thus, FA patients have a more than 700-fold higher HNSCC incidence than the normal population [55,56]. Based on reports in the literature, it was estimated that FA-patients also have a 2000-fold higher risk of developing esophageal cancer [2]. Although more FA patients are now surviving the first two decades of life, over-all survival of FA patients with HNSCC is poor due to a high percentage of tumor progress under therapy and also early relapses in combination with severe toxicities of the usually reduced irradiation and/or chemotherapy [2,7,34,63].

Behavioral risk factors for HNSCC

Behavioral consumption of alcohol and smoking/tobacco use are well established risk factors for the development of HNSCC [44,54]. Estimates suggest that in sporadic HNSCC, heavy smoking increases the tumor risk 20-fold and heavy alcohol consumption 5-fold [54]. In combination, both factors lead to a 50-fold elevated risk for the development of HNSCC [54] and also negatively impact survival [44]. Additional factors that promote the development of HNSCC are eating habits with low proportions of fruits and vegetables [39], bad dental hygiene [57], and chronic exposure to betel or areca nuts [44,54].

In contrast to the normal population, the majority of FA patients experience HNSCC frequently at a very young age and often without exposure to any of these risk factors. Only 3 out of 19 patients with HNSCC registered in the IFAR [35] and only one third (4 out of 12) of FA patients in a follow-up study indicated the consumption of nicotine and/or alcohol [7].

HPV/p53

In 1983, infection with oncogenic human papilloma viruses (HPVs) was first described in the context of sporadic HNSCC[68]. Today, local infections with oncogenic HPV are well established to play a major role in the pathogenesis of sporadic HNSCC [19,51,71] and other human cancers [47,66]. From more than 120 different HPV subtypes that have been isolated, infections with the high-risk strains HPV 16 (approximately 90% of cases) and HPV 18 are prevalent in patients with epithelial cancers world-wide [19,47,51]. The transforming activity of these high-risk serotypes is predominantly the consequence of two viral oncogenes, E6 and E7, that are expressed in HPV-infected cells (reviewed in [46]). E6 binds to the DNA-binding region of human P53 and therefore inhibits its tumor suppressive functions, including induction of apoptosis and cell cycle arrest. E7 mediates poly-ubiquitination and thereby degradation of the tumor suppressor protein RB [46]. E7 also inhibits the tumor suppressor p21 (CIP1) and indirectly leads to reduced cell cycle arrest and thus to increased cell proliferation. Both viral oncogenes are also involved in the activation of the WNT signaling pathway, at least in HPV16 positive gynecological SCCs [46].

Multiple studies have demonstrated that HPV infections in the normal population are predominantly transmitted through sexual activities [13,14,19,30]. Especially certain sexual behaviors such as open-mouthed kissing, oral sex, oral-anal contacts and a high number of sexual partners have been strongly associated with an elevated risk of developing an HPV positive tumor [13,14,30]. However, it is important to note that it is completely unknown how and when HPV infection occurs in FA patients. Infection with HPV16 virus activates the FA pathway in normal cells and increases the genomic instability in FA cells [64,65]. A genetic cross showed that Fancd2 deficient mice transgenic for the HPV16 E7 oncogene have a higher incidence of chemically induced HNSCC compared to the Fancd2 deficient control animals[49]. In FA patients however, there is conflicting data on the impact of HPV for the pathogenesis of SCCs. Kutler et al. detected HPV DNA in 84% of SCC samples from 25 FA patients, compared to only 36% of specimens from their nonFA control tumor group [34]. 15 of the 18 HNSCCs and 6 of the 7 of the anogential SCC were positive in their FA patient group [34]. Since the HPV E6 protein inhibits p53 [59,60], they also sequenced p53 in tumor tissues from both groups. No mutations in p53 were detected in samples from FA patients compared to 36% of tumors in their control group [36]. These observations in FA patients and the findings in mice [49] strongly support the hypothesis that defects in the FA/BRCA pathway are associated with increased susceptibility to HPV infections and therefore a higher propensity for developing HPV-triggered SCCs [36].

In sharp contrast to this, van Zeeburg et al. could not detect HPV DNA in 4 HNSCC cell lines from FA patients or in 7 HNSCC cell lines from patients with sporadic tumors [73]. In addition, p53 mutations were detected in all FA HNSCC cell lines and in 4 of the 7 cell lines from sporadic HNSCC patients [72,73]. In a follow-up publication, van Zeeburg et al. again did not detect HPV DNA in 16 HNSCC specimens from FA patients and indirect HPV analysis by p16 immunostaining only showed positive staining in tumors from 2 of the 13 analyzed FA patients [74]. P53 mutations were detected in the majority (8 out of 13) of analyzed patients [74].

These completely opposite findings obtained from FA patients in the U.S.A. and in Europe currently makes it impossible to finally judge the impact of HPV infections on the highly increased incidence of SSCs in FA patients. However, it appears worth mentioning that the incidence of HPV infection and p53 mutations in nonFA patients reported by Kutler et al. in their control group of sporadic SCCs [34] is similar to what has been reported by others in independent studies [19,20]. In addition, other authors have also reported high percentages of HPV infection in their FA patients [15,23]. Our own findings revealed WV infection in only one out of 30 sporadic HNSCC cell lines, whereas analysis of 123 tumor specimens detected HPV infection in 37 cases (30%) [4,24].

Therapy of HNSCC

The current state-of-the-art treatment of HNSCC is a multidisciplinary approach, combining surgery, chemo- and radiotherapy options [6,61]. Primary surgical removal of the cancers with clear tumor cell-free margins followed by local radiotherapy and concurrent delivery of high-dose chemotherapy [17] has improved the 5-year survival rate in the majority of HNSCC patients significantly 16,37,61]. The backbone of the chemotherapy regimen is the crosslinking alkylating agent cisplatin which is usually given in intermediate high-doses of 50–100 mg/m² [6,61]. Addition of other drugs such as 5-fluorouracil, methotrexate, and paclitaxel may increase the prognosis in some instances, but also have been associated with increased toxicities that limited their clinical utility [43].

Unfortunately, all cells of FA patients are exquisitely sensitive to DNA cross-linking agents [16,31]. Therefore, systemic cisplatin treatment cannot be used in this patient population due to the very high toxicities and organ failures induced even by reduced dosing leading to fatal outcome [7,63]. Consequently, radiotherapy can be used as a localized cancer treatment approach in FA [26], but it also seems to be associated with severe complications in FA HNSCC patients [7,41,69]. A recent larger study confirmed that radiotherapy is problematic in FA patients, as one third of FA patients (4 out of 12) died during the course of therapy. Pancytopenia was observed in half of the patients and most of them suffered from partly severe mucositis and dysphagia [7]. In 2002, the outcome for 14 FA patients with SCCs who received radiotherapy was summarized from the literature [1]. Cancers in this cohort included 10 HN, 3 esophageal and 1 vaginal SSCs. Although the numbers are small, it appears as if local irradiation >34 Gray was associated with severe toxicities, mucositis, edema, ulceration or local bleeding in the 10 HNSCC patients, and only two patients in the literature were alive 3 and 10 months after the initial diagnosis of the cancer [1]. Unfortunately, these severe clinical complications of the toxic radiotherapy are not necessarily predictable by in vitro sensitivity testing of lymphoid cells of FA patients [41]. As a consequence of the high iatrogenic morbidity and toxicity, individualized therapy approaches were suggested for FA patients with extended courses at lower daily doses (150–180 cGray per fraction) and intensified monitoring of the hematological and organ toxicities during treatment [7].

Based on these experiences, it is obvious that the most important therapeutic option in FA patients with HNSCC is the complete surgical resection of the cancer at an early stage (Fig. 1). This should be possible in the majority of FA tumor patients, as two thirds of tumors appear to be located in the oral cavity [35]. Surgery can either be performed conventionally or in certain cases by laser resection (Scheckenbach et al. unpublished). Neck dissections should be added if any lymph node metastasis appears even remotely possible, as loco-regional tumor control by other means might be more difficult to achieve [7].

Fig. 1.

Algorithm for prophylaxis and treatment in HNSCCs of FA patients (For details, see text).

In sporadic HNSCC, targeted therapies have been a major focus of efforts to increase the dismal prognosis for advanced stages in the last years [43]. The most developed approach here is based on the fact that the Epidermal Growth Factor Receptor (EGFR) is often overexpressed and activated in sporadic HNSCC [9]. As the EGFR is a proto-oncogene, an increased activity promotes tumor progression and invasion, inhibition of apoptosis, angiogenesis and metastasis [9]. Importantly, although the EGFR targeting antibody Cetuximab is FDA-approved in combination with radiotherapy for advanced HNSCC [9], therapy with EGFR antibodies is not a sufficient alternative to radiation or chemotherapy with cisplatin in sporadic HNSCC [62]. In addition, treatment with Cetuximab is associated with a wide spectrum of adverse reactions [38] and instances of using Cetuximab in FA patients have not been reported so far.

Recently, photodynamic therapy (PDT) based on an interaction between oxygen, light and a photosensitizer to induce apoptosis/necrosis has been introduced for local tumor control [28]. Experiences in FA - neither in FA cells in vitro nor in patients in vivo - have not been reported. Finally, based on the fact that p53 is either mutated or inactivated by HPV infections in the majority of HNSCCs, clinical trials have been conducted with adenoviruses specifically targeting cells with nonfunctional p53 [38]. None of these approaches has received FDA approval yet and would thus be readily available for FA patients with HNSCCs.

Therefore, due to our very limited treatment options for HNSCC in FA patients, it appears mandatory to systematically collect and analyze the treatments and outcomes that have been observed in FA patients worldwide. As most cases of HNSCCs in FA patients have never been published, a feasible approach might be to ask the national FA family support groups for access to FA patients and their families and thereby expand our knowledge for what has been tried and achieved in FA patient clinical care so far on an individualized basis.

Screening and prevention of HNSCC

The extremely high incidence of HNSCC and the uttermost importance of early surgical interventions to achieve cure for HNSCC in FA patients emphasizes the need for regular and rigorous surveillance measures. In addition, as only two thirds of all HNSCCs in FA patients are located in the oral cavity [35], surveillance should ideally be performed by a specialist (e.g. ENT or oral surgeon) and should also include the naso-, oro- and hypopharynx as well as the larynx and possibly the esophagus, especially in older patients and if there are any signs of reflux or dysphagia. Semiannual examinations might already be indicated as early as 10–12 years of age, particularly if the patient had undergone stem cell transplantation [42,53]. Without prior transplantation, screening could start later at the age of 15 years, however extensive examinations of the upper aerodigestive tract every 6–8 weeks appears necessary in FA patients with leukoplakia and recurrent oral lesions [34]. To thoroughly access all the different areas where HNSCC can develop, the use of a flexible endoscope is inevitable (Fig. 1).

Early detection of pre-malignant lesions or HNSCC can be sought by a variety of screening methods [45]. Different light-based screening aids are in development based on the fact that abnormal metabolic and structural changes lead to differences in absorbance and reflectance properties [32]. Analogous to the detection of cervical cancer, brush biopsies can be performed as a milder alternative to scalpel biopsies [33]. Since no tissue specimen but only single cells are available in brush biopsies, these probes are limited to cytological investigations; however, additional DNA ploidy analysis can complement these cytological investigations [33]. Several other test methods are in development and studied in different centers [45].

Increased surveillance should be performed for any dysplastic lesions. In general, even mildly dysplastic areas should be removed if feasible (Fig. 1). Finally, essential for the prevention of HNSCC is avoidance of additional risk factors such as nicotine and alcohol, maintenance of a healthy lifestyle and careful oral hygiene. In a transplant setting, it appears to be very important to avoid acute or chronic GvHD≥II, especially in combination with irradiation for conditioning.

Finally, although the impact of HPV for the development of SCC in FA patients has not been clarified yet, standard HPV vaccination for girls and women can be considered a safe procedure, based on experiences with more than 40 million doses of vaccines distributed in the U.S.A. alone (Center for Disease Control and Prevention (CDC); October 25, 2011: http://www.cdc.gov/media/releases/2011/t1025_hpv_12yroldvaccine.html). Vaccination seems to have the greatest impact on antibody titer and protection when given between 11–12 years of age and before any exposure to HPV. There are two different vaccines available, the quadrivalent Gardasil® from Merck (http://www.ema.europa.eu/ema/index.jsp?curl=pages/medicines/human/medicines/000703/human_ med_000805.sjsp&jsenabled=false) and the bivalent Cervarix® from GlaxoSmithKline (http://www.ema.europa.eu/ema/index.jsp?curl=pages/medicines/human/medicines/000721/human_med_000694.sjsp&jsenabled=false), both using recombinant virus-like particles without any viral DNA. While Gardasil is effective against the oncogenic HPV16 and −18 and also the low risk HPV6 and −11 strains associated genital warts, Cervarix only includes HPV16 and −18, but may induce higher antibody titers after standard vaccination [8,22]. In 2009, the quadrivalent HPV vaccine was also approved by the FDA for usage in boys and men and is thought to prevent genital warts and anal cancer, which in more than 80% are associated with HPV. In October 2011, the Advisory Committee on Immunization Practices (ACIP) in the U.S.A. recommended to also vaccinate all 11- and 12-year-old males against HPV using the quadrivalent vaccine. Although this will be associated with significant costs, the expert panel considered the vaccination an important opportunity to reduce the spread of HPV from males to females and to decrease the burden of HPV-related diseases in both genders (http://www.cdc.gov/media/releases/2011/t1025_hpv_12yroldvaccine.html).

Therefore, based on the exquisite susceptibility of FA cells to HPV infections [49,64,65], the excellent safety profiles of the available vaccine(s) and the general recommendation for vaccination of all 10-to 11-year-old children against HPV in the USA, we consider it important to now include both genders of FA patients in the vaccination approaches for HPV16 and −18 and perhaps also HPV6 and −11 worldwide. If future studies reveal that FA patients have a greatly increased susceptibility, possibly in combination with an immunological defect to clear HPV, it might even be appropriate to vaccinate FA patients for HPV immediately after diagnosis of FA.

Conclusion

HNSCCs are a growing problem in the clinical care of FA patients, due to both, the high incidence in transplanted but also in non-transplanted patients and the very limited therapeutic options for manifest malignancies. Early detection and surgical removal are paramount to overcome the poor outcome of HNSCCs in FA patients. Prospectively, vaccination for oncogenic HPV may reduce the occurrence of HNSCC and new conservative therapeutic strategies that are not based on cross-linking chemotherapeutics and irradiation such as targeted therapies may improve survival of FA patients with HNSCC. Nevertheless, ongoing and future efforts should also include standardized diagnostic and therapeutic treatments for all FA patients and an intensified exchange of medical and genetic information between different clinics and institutions worldwide. Since FA patients with HNSCC are such a small and difficult-to-achieve-cure collective, it is absolutely necessary to share experiences and systematically develop and standardize new prevention and treatment approaches. This can only be achieved by a unified approach.