Treatment of Fanconi Anemia–Associated Head and Neck Cancer: Opportunities to Improve Outcomes

Fanconi anemia is an inherited disease of genomic instability characterized by progressive bone marrow failure, congenital growth defects, and increased predisposition to cancer. Although patients with Fanconi anemia are at high risk for many hematologic and solid malignancies, the proclivity for mucosal squamous cell carcinomas (SCC), especially head and neck SCC (HNSCC), is particularly striking. Fanconi anemia–associated HNSCC (FA-HNSCC) occurs with an estimated incidence that is 500-fold to 800-fold greater than HNSCC in the general population, frequently developing at young ages with generally poor survival (1, 2). With the substantial advances in hematologic management of Fanconi anemia and the associated improvement in life expectancy for these patients, the number of patients at risk to develop FA-HNSCC will likely increase.

Patients with Fanconi anemia demonstrate an intrinsic sensitivity to DNA-damaging agents including chemotherapy and radiotherapy, at least in part because the FA-BRCA pathway plays a critical role in DNA repair (3). This poses unique challenges to the treatment of FA-HNSCC because chemoradiation is a standard approach for pharyngeal and laryngeal cancers and adjuvant radiation or chemoradiation is commonly employed following surgical resection of oral cavity cancers with cervical lymph node involvement. Although surgery is generally well tolerated in this population, cytotoxic chemotherapy and radiotherapy are less commonly utilized because of legitimate concerns of severe toxicity (4). There are few reports of administration of the FDA-approved agents cetuximab, pembrolizumab, and/or nivolumab to patients with FA-HNSCC and the efficacy of these agents in treating HNSCC that arises in the setting of Fanconi anemia remains unclear. To date, there is a paucity of evidence in the literature to guide therapy of FA-HNSCC, and the details of existing individual case reports have not been comprehensively synthesized to inform treatment decisions for future patients with FA-HNSCC. In this systematic review, we identified all reported cases of FA-HNSCC in the peer-reviewed literature that described both the treatment regimen and patient outcome from 1966 to 2020, including treatment-related toxicity and survival. We now summarize these cases with a focus on clinical responses to the regimen(s) employed and treatment-associated toxicities.

Worldwide, parents of children with Fanconi anemia gravitate toward academic centers of excellence which, over time, have accumulated substantial expertise in managing their child's bone marrow failure. However, because Fanconi anemia is a rare genetic disorder and mucosal SCCs, including HNSCCs, tend to arise years later in adulthood, the multidisciplinary cancer-oriented teams who manage these malignancies rarely have any experience with patients with Fanconi anemia. Surgical resection remains the primary treatment modality for primary oral cavity cancers and this is no different for FA-HNSCC. But few centers have administered systemic therapy and/or radiation as primary or adjuvant therapy to these individuals. To our knowledge, there are no reports of clinical trials dedicated to FA-HNSCC and no ongoing therapeutic studies, underscoring the challenges of developing evidence-based treatment approaches. This review is intended to serve as a guide for clinicians who encounter a patient with Fanconi anemia with HNSCC.

Fanconi anemia arises from predominantly biallelic germline mutations in any of 22 identified genes of the Fanconi anemia complementation group: FANCA, FANCB, FANCC, FANCD1 (BRCA2), FANCD2, FANCE, FANCF, FANCG, FANCI, FANCJ (BRIP1), FANCL, FANCM, FANCN (PALB2), FANCO (RAD51C), FANCP (SLX4), FANCQ (ERCC4/XPF), FANCR (RAD51), FANCS (BRCA1), FANCT (UBE2T), FANCU (XRCC2), FANCV (MAD2L2), or FANCW (RFWD3) (5). Broadly, gene products of the FANC family function in concert to orchestrate repair of DNA interstrand crosslinks (ICL), lesions which halt DNA synthesis and transcription (3, 6). Notably, the identification of multiple FANC genes as prominent susceptibility loci for breast cancer (FANCD1 as BRCA2 and FANCS as BRCA1) resulted in the naming of this integrated repair pathway FA-BRCA (7). Recognition of stalled replication forks triggers recruitment of the Fanconi anemia core complex, a multimeric ubiquitin ligase comprised of at least eight FANC protein subunits (8–10). Association of the core complex with an E2-conjugating ligase allows for monoubiquitination of a FANCD2-I heterodimer, a key step that initiates downstream formation of the ICL repairosome (11, 12). Endonucleases are subsequently recruited to cleave DNA at the ICL site, after which translesion synthesis and homologous recombination occur to restore the original DNA duplexes (13–15). This process is illustrated in greater detail in Fig. 1.

Defective ICL repair and genomic instability underlie the central pathophysiology of Fanconi anemia. Without the ability to repair ICLs, cells are prone to DNA breakage and rearrangement, increasing the risk for oncogene alteration and loss of tumor suppressor genes (3). In addition, deficiencies in FANC proteins lead to further cellular dysregulation resulting in excess cytokine production, inflammasome activation, cell-cycle defects, and increased sensitivity to free radicals and aldehydes (16). This constellation of genetic and cellular alterations is thought to contribute to the high incidence of hematologic disorders and markedly elevated cancer predisposition in individuals with Fanconi anemia.

Although Fanconi anemia is rare, with an incidence of approximately 1 in 130,000 births, it is the most frequent genetic cause of bone marrow failure (17–20). The carrier frequency is estimated at 1:181 in the United States, but is notably higher in Israel with a frequency of approximately 1:93 (21). In addition to the increased frequency in Ashkenazi Jews, Fanconi anemia is also more common in Spanish Gypsies and Afrikaners (17, 21–24). For unknown reasons, Fanconi anemia preferentially affects males, with a male-to-female ratio of 1.2:1 (25). The majority of pathologic FANC variants are inherited in an autosomal recessive pattern with the exception of FANCB which causes an X-linked recessive form of Fanconi anemia, and FANCR which manifests as an autosomal dominant Fanconi anemia–like syndrome (26). The most commonly mutated gene in cases of Fanconi anemia is FANCA, responsible for 60% to 70% of cases, followed by FANCC (10%–15%), and FANCG (10%; refs. 26, 27).

Many patients with Fanconi anemia display characteristic congenital abnormalities, including short stature, skeletal malformations (especially of the upper extremities), abnormal skin pigmentation, decreased fertility, and renal anomalies (28). However, Fanconi anemia is a highly heterogenous disease, and different FANC mutations are associated with remarkably disparate clinical syndromes (29, 30). Mutations of various FANC genes result not only in diverse external phenotypes, but distinct propensities for the development of myelodysplasias and cancer (31, 32). Without medical management, patients with Fanconi anemia are at high risk of severe hematologic disorders including bone marrow failure, aplastic anemia, myelodysplastic syndrome, and acute myelogenous leukemia (33). The only treatment which can restore adequate hematopoiesis in Fanconi anemia remains hematopoietic stem cell transplantation (SCT; ref. 34). While early attempts (over 20 years ago) at SCT for Fanconi anemia were associated with a high mortality, recent advances in patient selection, HLA donor typing, and conditioning regimens have led to 5-year survival rates currently as high as 70% to 94% (35–38). These improvements in hematologic management have substantially increased the survival of patients with Fanconi anemia into adulthood. The management of solid tumors, specifically HNSCC, now represents the foremost threat to life for these patients with Fanconi anemia.

The overall risk of developing any solid tumor is approximately 40-fold greater in Fanconi anemia relative to individuals without Fanconi anemia, with a cumulative incidence exceeding 75% by age 45 (39–41). Occurrence of head and neck, esophageal, and vulvar SCCs in particular is substantially higher among patients with Fanconi anemia compared with the general population (40). The increased risk of developing HNSCC in individuals with Fanconi anemia (500- to 800-fold) is especially notable (1, 2). However, the mechanisms underlying this extreme predilection for developing SCCs at these particular mucosal sites are poorly understood (42). It remains conceivable that the increase in oral cavity cancer in young nonsmokers can be attributed to underrecognized germline mutations in FANC-related genes (43). Premalignant and malignant lesions affecting the lateral tongue and buccal mucosa often occur adjacent to dentition, supporting a possible relationship to dental trauma and metabolites of the microbiome, which emphasizes the importance of attention to oral examination and hygiene in this population.

Although the diagnosis of Fanconi anemia is made before the development of HNSCC in the majority of cases, the phenotypic heterogeneity of Fanconi anemia necessitates a high index of suspicion in all teenage or young adult patients who unexpectedly develop HNSCC. Strikingly, some patients with Fanconi anemia display no obvious physical manifestations of the disease (25, 44, 45). Thus, the presentation of Fanconi anemia–related cancer and susceptibility to DNA damage in the absence of bone marrow failure may be related to differences in effects of genetic alterations in Fanconi anemia pathway, modifiers, or mosaicism (30). We identified four case reports of young patients without an antecedent Fanconi anemia diagnosis, who were subsequently diagnosed with Fanconi anemia as adults based on toxic reactions to the HNSCC or esophageal SCC treatment regimen. These patients ranged in age from 24 to 51, and developed severe toxicity including sepsis from neutropenia with as little as a single dose of platinum chemotherapy administered with radiotherapy (46–49). In three of the four patients, the toxicity proved fatal. These sobering reports highlight the importance of considering Fanconi anemia in the differential diagnosis of young patients with HNSCC.

Exposure to tobacco and alcohol are well-characterized risk factors for developing head and neck cancers (50). However, despite declining tobacco usage, the overall incidence of oropharyngeal SCC (OPSCC) is increasing rapidly in the United States and Western Europe (51). This rise is largely attributed to human papillomavirus (HPV) infection, with a majority arising from high-risk strain HPV strains, especially HPV16 (52). In the United States, the incidence of HPV-positive OPSCC is now estimated to be 2.5-fold higher than that of HPV-negative OPSCC (53). To date, three studies have evaluated the HPV status of FA-HNSCC tumors using molecular methods. One group assayed 18 FA-HNSCC tumors (15 oral cavity, two pharynx, and one larynx) using PCR, finding that 15 of 18 FA-HNSCC cases were positive for DNA from HPV16 or HPV18 (54). However, these findings remain controversial. A subsequent study also utilizing PCR reported that none of the 16 FA-HNSCC tumors assessed had detectable HPV DNA (55). The third study, from 2013, also found no HPV DNA via PCR in the five FA-HNSCC cases analyzed (three oral cavity, two oropharynx; ref. 56). Detailed genomic analysis of FA-HNSCC will be necessary to further delineate the role of HPV in the pathogenesis of these cancers.

The anatomic subsite distribution of FA-HNSCC is consistent with HPV-negative HNSCC. In the general population, HPV-positive HNSCC occurs most commonly in the oropharynx, particularly affecting the lingual and palatine tonsils (57). FA-HNSCC generally occurs in the oral cavity, at subsites typically associated with HPV-negative disease (58, 59). The low incidence of oropharyngeal cancer relative to oral cavity tumors in patients with Fanconi anemia suggests that FA-HNSCC is driven by lack of proper repair of DNA lesions, rather than increased HPV-mediated carcinogenesis (60). Nevertheless, it is essential for patients with Fanconi anemia to receive preventive HPV vaccination, just as it is for any young adult. There are currently three HPV vaccines approved by the FDA in various demographics: bivalent, quadrivalent, and nonavalent, with the latter indicated for both males and females ages 9 to 45. Because the majority of vulvar (and vaginal) SCCs arising in patients with Fanconi anemia are associated with HPV, all patients with Fanconi anemia should receive the HPV vaccine as children, prior to engaging in sexual activity and, hence, HPV exposure.

The median age of diagnosis for sporadic HNSCC is 66 for HPV-negative cancers and 53 for HPV-positive cancers (61). In individuals with Fanconi anemia, HNSCC develops much earlier, with a median onset of approximately 30 years of age (1, 4, 58). Patients with Fanconi anemia who have undergone SCT present with HNSCC at even younger ages than untransplanted individuals with a median age of 18 to 20, possibly due to the carcinogenic effect of cytotoxic conditioning regimens employed in preparation for SCT, initiating and promoting effects of inflammation by GVHD, or decreased cancer immune surveillance (59, 62). Notably, HNSCC (and specifically oropharyngeal cancer) is also the most frequent secondary malignancy among patients without Fanconi anemia who receive SCT, developing at a seven to 16 times greater incidence than is seen in the general population (63). The oral cavity is the most common site of FA-HNSCC, with approximately 60% of Fanconi anemia–associated oral cavity carcinomas affecting the oral tongue (58, 59). FA-HNSCC frequently presents at advanced stages with correspondingly poor survival, which underscores the need for active surveillance including frequent oral cavity exams in all patients with Fanconi anemia and definitive management of high-risk premalignant lesions. Early detection of cancer allows for treatment with surgery alone, conferring the highest chance of survival, as well as the potential to avoid radiation and chemotherapy treatments. However, the largest cohort study to date, utilizing the International Fanconi Anemia Registry, found that of 35 FA-HNSCC cases, 15 (43%) presented as stage IV, of which two-thirds of these stage IV cases (10/15) had N2b disease or greater, indicating advanced cancer spread to the neck. Furthermore, after surgical resection, definitive radiotherapy, and/or adjuvant radiotherapy/chemoradiation, recurrence of FA-HNSCC occurred in 48% (17/35) of patients with a 5-year Kaplan–Meier survival estimate of only 39% (4).

Multiple factors complicate developing a treatment plan for FA-HNSCC. Individuals with Fanconi anemia are likely to experience toxicity when treated with agents which cross-link DNA, generating great concern over the use of many of the FDA-approved chemotherapy agents (notably platinum-based compounds) as well as radiotherapy. As a result, patients with FA-HNSCC often undergo ablative surgeries with the goal of completely resecting locoregional disease, without adjuvant radiotherapy or chemoradiation. Surgery itself is generally well tolerated, but late presentation and a reluctance to administer adjuvant chemotherapy or radiotherapy (or chemoradiation) in FA-HNSCC results in discouragingly high rates of recurrence (4, 64, 65). In contrast to surgical management, the efficacy and tolerability of chemotherapy or radiotherapy in patients with FA-HNSCC remains unclear. Here, we review the cases of FA-HNSCC treated with surgery, radiotherapy, and/or systemic agents and summarize the associated toxicities and clinical outcomes.

To define the review cohort, a PubMed search was performed using the terms: “(Fanconi anemia) AND (oral OR head OR neck OR mouth OR tongue OR buccal OR pharynx OR pharyngeal OR oropharynx OR oropharyngeal OR larynx OR laryngeal OR esophagus OR esophageal) AND (carcinoma OR cancer).” Reports which pertained to Fanconi anemia–associated SCCs of the oral cavity, oropharynx, larynx, or esophagus were selected and aggregated for analysis without restrictions on publication date, which ranged from February 1966 to December 2020. All patients who were treated with surgical resection, radiotherapy, systemic therapy, or a combination of modalities were included. Overall, we identified a total of 119 patients with FA-HNSCC (including 16 with esophageal SCC) with these criteria, with demographic and clinical information summarized in Table 1. The median age of diagnosis at first primary was 28 years old, which is consistent with the reported onset of FA-HNSCC in prior epidemiologic studies (1, 58). The cohort was 52% (62/119) female, an intriguing observation given that both Fanconi anemia and HNSCC are, in general, more common in men than women (25). Approximately half (58/119) of the patients received a bone marrow transplantation prior to their HNSCC diagnosis. Of the 47 patients with available complementation group information, the two most common mutations were FANCA (68%) and FANCC (17%), though mutational data was not reported for 60% (72/119) of the cohort.

The 119 individual patients identified developed a total of 131 separate primary HNSCC tumors, with 11 patients having multiple distinct primaries of the head, neck, and/or esophagus (synchronous in one patient and metachronous in 10 patients, with one patient developing three metachronous oral cavity tumors). Of these primaries, 69% (91/131) were oral cavity tumors. Eighty-four primaries had staging data, of which 39% (33/84) were stage IV. Twenty-seven tumors were assessed for HPV status, of which 11 (41%) were HPV negative (one oropharynx, 10 non-oropharynx) and 16 (59%) were deemed HPV positive (two oropharynx, 14 non-oropharynx). However, it is important to note that 15 of 16 HPV-positive tumors were reported from the same controversial study described above, which reported all HPV-positive samples as HPV16 with a very high number in non-oropharyngeal locations (13 were oral cavity primaries; refs. 4, 54).

Of 131 primary tumors, 106 were treated with primary surgical resection. Of these, 22 cases were received adjuvant radiotherapy, eight had adjuvant chemoradiation, one was treated with neoadjuvant chemoradiation, and one received adjuvant chemotherapy therapy alone. When the primary disease was deemed “unresectable,” nine tumors were treated with primary radiotherapy, nine with primary CRT, and six with chemotherapy only. One patient underwent salvage surgery only after prematurely terminating chemotherapy due to toxicity. These patients and their outcomes are summarized in Tables 2 through 7. Of the 119 total patients, 43 were reported to experience at least one recurrence of their disease (distinct from the 12 second primary tumors). Given that patients with Fanconi anemia generally tolerate surgery well, FA-HNSCC primaries treated with surgical resection alone will not be discussed further. There are 26 reports detailing one or more of these cases (4, 42, 48, 64, 66–87).

Surgical resection is the preferred primary treatment modality for oral cavity tumors, and carries a high rate of cure for early-stage disease (88, 89). Unilateral or bilateral neck dissections are often performed to allow for more accurate pathologic staging and to remove nodal metastases (90). Adverse pathologic features (such as extranodal extension, positive resection margins, multiple involved nodes, and lymphovascular or perineural invasion) inform the decision to administer adjuvant treatment. In patients who have undergone surgery with curative intent, advanced lymph node involvement or the presence of the aforementioned adverse pathologic features typically warrant a course of adjuvant radiotherapy, or in rare cases of highly concerning features, adjuvant chemoradiation. In contrast to oral cavity tumors, cancers of the oropharynx, larynx, and hypopharynx are commonly treated with definitive chemoradiation as primary therapy, particularly in cases where surgery would result in unacceptable functional deficit (90).

When radiotherapy is employed, intensity-modulated radiotherapy (IMRT) is the current standard of care in HNSCC, and demonstrates an improved toxicity profile with better locoregional control compared with historical three-dimensional conformal radiotherapy (91). Completing a course of radiotherapy, typically administered 5 days a week over the course of 6 to 7 weeks without interruption, is key for maximal efficacy, as breaks in radiotherapy are associated with poorer outcomes (92, 93). However, radiation toxicities of varying degrees are common among all patients with HNSCC, including dermatitis, edema, mucositis, pain, dysphagia, and xerostomia (94). Dose-limiting side effects can easily interfere with the delivery of uninterrupted radiotherapy, especially in patients with Fanconi anemia who are intrinsically sensitive to DNA damage. Thus, planning for enhanced supportive care, with ability to adapt the treatment in case of poor tolerance, must be a part of the pretreatment discussion. A key limitation in reviewing the evidence of radiotherapy tolerability in FA-HNSCC is the lack of detail specifying the precise radiation modality utilized in the majority of patients, as reports of radiotherapy-treated FA-HNSCC date back to the 1960s (95).

Initial use of IMRT began in the 1990s and became the standard of care for HNSCC with widespread adoption in the 2000s, with efforts to standardize IMRT across institutions in the 2010s. A major advantage of IMRT is enabling initial fields to encompass the lesion and monitor toxicity and response, before widening the field to provide lower cumulative doses planned for adjacent mucosa or neck. The use of IMRT is mentioned specifically in only three patient cases (65, 96, 97), but is presumed to be the current standard approach for delivery of external beam radiotherapy for all HNSCC, including FA-HNSCC. Proton beam radiotherapy has been reported in a single case treated with 70.4 CGE (cobalt Gray equivalent) to the high-risk volume and 60 CGE to the remaining ipsilateral neck, with concurrent cetuximab. The patient tolerated the regimen but developed an in-field recurrence 16 months after completing therapy (65). Proton therapy is highly intriguing due to its increased conformality to targeted areas and is frequently used for these reasons in pediatric and young adult radiation treatments. However, because of the potential for less dose sparing of the superficial skin and mucosal surfaces, the advantages of this radiation modality must be more clearly established in patients with FA-HNSCC before it can be recommended as a rule.

Postoperative radiotherapy is typically administered once daily, five times per week at 1.8 to 2 Gy as conventionally fractionated irradiation; a typical goal is to achieve a total dose of 60 to 66 Gy over the course of 6–6.5 weeks (RTOG 9501 and EORTC 22931 trials; refs. 98, 99). Twenty-two patients with FA-HNSCC (21 HNSCC, one esophagus) underwent surgery followed by adjuvant radiotherapy for primary tumors (Table 2). Radiotherapy dosage was reported for 15 of 22 patients, who achieved a median total dose of 51.8 Gy. Eleven of 22 patients completed the planned radiotherapy course, while six of 22 had their treatment interrupted or terminated because of radiotherapy toxicities. Completion status was not available for five patients. Of the 11 patients who completed adjuvant radiotherapy, eight died of disease without further treatment, one underwent subsequent surgery and adjuvant chemotherapy for recurrence (Nolan 2017; ref. 100), one was alive 19 months postoperatively without recurrence of her initial tumor but with a new primary (Kaplan 2011; ref. 87), and one was alive with no evidence of disease (NED) over 14 years postoperatively (Kutler patient No. 30; ref. 4). The six patients who interrupted or terminated radiotherapy early all succumbed to disease within 20 months of their initial surgery. Of the remaining five patients without radiotherapy completion status, three died of disease, one was still undergoing treatment, while 1 was alive with cancer progression at time of publication and presumably succumbed to disease. It should be noted that patients for whom postoperative radiotherapy is recommended are those with more advanced cancers, and the outcomes are expected to be worse than the surgery-alone population.

Definitive, or curative-intent, radiotherapy is typically administered once daily, five times per week at 2 Gy as conventionally fractionated irradiation aiming to achieve a total dose of 70 Gy over the course of 7 weeks. Nine patients (seven HNSCC, two esophagus) received radiotherapy without surgery for primary disease, 56% of whom (5/9) completed the course and 44% of whom terminated treatment early due to toxicity (1 during radiotherapy; Table 3). The average total dose achieved was 52 Gy. Of the 5 who completed primary radiotherapy, four died of disease (one from a separate primary) and one was alive with recurrence 14 months after diagnosis (Budrukkar 2009; ref. 101). Of the four patients who terminated radiotherapy, three succumbed to disease, while one was alive with disease 2 months post-radiotherapy completion (Marcou 2001; ref. 102). Presumably, all patients alive with disease at the time of case report publication ultimately died of HNSCC.

In 4 patients, recurrent disease was treated with radiotherapy alone (Table 7). All died within 1 year of recurrence, with premature radiotherapy termination in two and unavailable completion information for one patient. The remaining patient completed a radiotherapy course of 32.5 Gy with palliative effect, although failure to control disease led to widespread metastasis and death (Snow 1991; ref. 74).

In summary, radiotherapy was relatively well tolerated in patients with FA-HNSCC both in the definitive and adjuvant treatment setting. For recurrent FA-HNSCC, radiotherapy was less well tolerated with only one of three patients completing the planned course. Overall, about 60% of patients successfully completed radiotherapy with an average total dose of 59.2 Gy, whereas 40% of patients who had to terminate early received total average dose of 31.8 Gy. Although the number of cases reported are relatively small, there is no evidence that prior SCT impacts radiation tolerability or toxicity.

Chemoradiation is indicated as the primary treatment for stage III/IV HNSCC where surgery is not feasible or as adjuvant therapy after surgical resection of tumors demonstrating pathologic high-risk features, such as positive margins and/or extranodal extension. The standard chemotherapy agent used for chemoradiation is cisplatin. Although cetuximab has been used in definitive chemoradiation, its use has declined after a number of recent clinical trials demonstrated its inferiority compared with cisplatin in HPV-positive HNSCC (103, 104). Patients described in this section received only conventional cytotoxic chemotherapy with or without radiotherapy. Patients who received molecular targeted agents with or without conventional chemotherapy or radiotherapy are presented in a subsequent section.

Five patients received chemoradiation with conventional cytotoxic chemotherapy in conjunction with surgical resection of primary tumors (Table 4; rows 1–5). This included four patients with HNSCC receiving adjuvant treatment, and one patient with esophageal carcinoma who underwent neoadjuvant chemoradiation. Agents administered included cisplatin and 5-fluorouracil (5-FU; one patient), cisplatin, bleomycin and methotrexate (one patient), methotrexate single agent (one patient), bleomycin single agent (one patient) for HNSCC and cisplatin and 5-FU for esophageal SCC. Of these five patients, one completed the chemoradiation course, threee terminated treatment due to toxicity, and completion status was not available for one patient. Four patients died of disease from 3 to 33 months following therapy. Only one patient was alive without disease at the time of publication (Hosoya 2010; ref. 66). Although chemoradiation was terminated early due to toxicity, her primary esophageal SCC “showed an excellent response to the preoperative chemoradiotherapy (66).” She developed a second primary of the oral cavity 5 years later, which was cured with resection, and remarkably, was alive with no evidence of either cancer at publication (6 years following esophagectomy, and 1 year after oral cavity surgery).

An additional seven patients with primary tumors (six HNSCC, one esophageal) received primary chemoradiation using conventional cytotoxic chemotherapy (cisplatin, carboplatin, gemcitabine, and/or 5-FU), described in Table 5 (rows 1–8). These seven patients account for eight total tumors, as one patient had multiple primaries treated with primary chemoradiation. Of the eight tumors treated with cytotoxic chemoradiation, only two treatment courses, involving non-platinum–based concurrent chemotherapy regimens with single agent 5-FU or single agent gemcitabine, were completed without interruption or termination (1/8 had completion status not available). All seven patients receiving chemoradiation died of disease, but the clinical course of one patient (Bremer patient No. 1; ref. 49) is particularly notable. This patient first presented with a base of tongue tumor at age 24, for which he received chemoradiation with carboplatin (60 mg/m²). However, carboplatin was discontinued after just one cycle due to pancytopenia. He completed 67.0 Gy radiotherapy total with one interruption at 38.4 Gy due to mucositis, after which he underwent surgical salvage (neck dissection) for persistent enlarged cervical lymph nodes. Remarkably, the patient remained free of disease for 5 years, at which time he developed two new primaries, of the oral cavity and anal canal. He received chemoradiation with gemcitabine (100 mg/m² and 25.2 Gy radiotherapy to the oral cavity, both of which were completed and well tolerated, but the patient succumbed to tumor progression 3 months later.

One patient with recurrent disease received conventional chemoradiation with agents and doses NA, but chemotherapy was terminated early due to “severe toxicity,” and the patient developed a second recurrence before dying of disease (Table 7; ref. 64).

Six patients (three HNSCC, three esophageal) received conventional chemotherapy alone without concurrent radiotherapy for primary tumors (five as intended primary therapy and one as adjuvant treatment), detailed in Table 6 (rows 1–6). Regimens administered included cisplatin and 5-FU, carboplatin and paclitaxel, cisplatin and 5-FU, 5-FU and cis-retinoic acid, and pepleomycin, a bleomycin-like drug. All six patients receiving cytotoxic chemotherapy either terminated treatment early, or information about treatment completion was unavailable. All six died of disease within 18 months of diagnosis. One additional patient received cis-retinoic acid alone without cytotoxic agents and developed recurrence (Table 6; row 7; ref. 105). It is notable that in the three patients with toxicity information available, all developed pancytopenia, one additionally complicated by sepsis (received cisplatin and 5-FU), and another experiencing liver failure and C. difficile colitis after two cycles of carboplatin and paclitaxel (65, 106).

One patient with recurrent disease received conventional chemotherapy alone, initially with capecitabine and then with oxaliplatin and 5-FU, but terminated therapy early due to toxicity (acute pericarditis and diarrhea), and died of disease within 6 months (Table 7; ref. 48).

In summary, cytotoxic chemotherapy (usually with a platinum-based regimen) with or without radiation was not well tolerated in the majority of patients with FA-HNSCC. Only 30% of patients who received primary chemoradiation were able to complete the planned treatment course. All patients who received cytotoxic chemotherapy alone (n = 7) had to terminate treatment early and died of disease.

Cetuximab was FDA approved in 2006 for the treatment of primary HNSCC in combination with radiotherapy and subsequently approved for recurrent or metastatic disease in combination with chemotherapy (the EXTREME regimen). In contrast, EGFR tyrosine kinase inhibitors (TKIs) have failed to improve outcomes in unselected HNSCC populations. Ten patients with FA-HNSCC received EGFR inhibitors (nine cetuximab, one gefitinib) either alone or in combination with other agents for treatment of primary or recurrent FA-HNSCC. Four patients (two oral cavity tumors, one pharynx cancer, and one larynx cancer) received adjuvant radiotherapy with concurrent cetuximab after surgical resection of primary disease, of which three completed the entire course of treatment with the standard toxicities of mucositis, dermatitis, dysphagia, and xerostomia using a mean radiotherapy dose of 70 Gy (Table 4; rows 6–9). All three patients who completed adjuvant cetuximab plus radiotherapy developed recurrences, one of whom died of disease without further treatment (Kutler patient No. 21; ref. 4), one who died following immunotherapy for recurrence (Beckham patient No. 8; ref. 65), and one who was alive with recurrence at 16 months post-chemoradiation of his primary and presumably succumbed to the cancer (Beckham patient No. 7; ref. 65). The patient who terminated adjuvant cetuximab and radiotherapy (total dose of 42.4 Gy) early experienced mucositis, dermatitis, and dysphagia along with cytopenia, wound breakdown and hemorrhage (Kutler patient No. 19; ref. 4). In addition, one patient (with primary oral cavity disease) received cetuximab with radiotherapy as primary treatment without surgery (Table 5; row 9). Although interruption or termination of chemoradiation was not reported, detailed toxicity information was not available, and he died of disease (Kutler patient No. 26).

Cetuximab was also administered to four patients with recurrent FA-HNSCC (Table 7). One patient (Beckham patient No. 5; ref. 65) received 10 fractions of radiotherapy with two doses of concurrent cetuximab before developing grade 3 mucositis and an abscess. After a break in therapy, he resumed radiotherapy without cetuximab, which was subsequently terminated at a cumulative dose of 42.4 Gy due to bleeding. Following radiotherapy termination, he began cetuximab therapy once again for 3 months with a “temporary” tumor response noted, before eventually dying of disease 7 months after radiotherapy cessation. Beckham patient No. 9 underwent adjuvant radiotherapy with two cycles of concurrent cetuximab, developing a carotid sentinel bleed before undergoing a third cycle with radiotherapy and paclitaxel. Following completion of chemoradiation, she continued cetuximab with two further doses of paclitaxel before receiving immunotherapy. While follow-up revealed a tumor response, she died shortly after administration of immunotherapy (discussed in the following section; ref. 65). Another patient received two cycles of chemotherapy consisting of cisplatin, 5-FU, and cetuximab (Nolan 2017; ref. 100). All systemic agents were terminated early after the patient developed sepsis and died. One patient received adjuvant radiotherapy with concurrent cetuximab for his recurrence (Wong 2013; ref. 86). He experienced neutropenia following a loading dose of cetuximab, and the cetuximab was discontinued for the final 2 weeks of radiotherapy (after eight total infusions) due to worsening radiation dermatitis. While cetuximab was terminated in this patient, he completed the planned radiotherapy course, with the authors reporting that both “radiotherapy and cetuximab were well tolerated with manageable toxicities (86).” The patient subsequently developed an extensive second recurrence and Pseudomonas bacteremia, dying 10 weeks after radiotherapy completion.

The European Medicines Agency approved the use of the EGFR TKIs gefitinib and afatinib for FA-HNSCC in 2018 based on compelling results in relevant preclinical models. There was one reported case of FA-HNSCC treatment with the small-molecule EGFR inhibitor gefitinib (Table 6; row 8). This patient, a 27-year-old female with tongue SCC, completed a 2-month course of gefitinib treatment without notable toxicity, including no rash or diarrhea (Jung 2005; ref. 107). Initially, her tumor size was substantially reduced (from 2.6 cm × 3.6 cm to 1.5 cm × 1.3 cm), but she later progressed on gefitinib. The patient subsequently received transhepatic chemoembolization for a hepatocellular carcinoma and died 2 months later of gastrointestinal bleeding.

In summary, concurrent cetuximab with radiation was well tolerated as primary treatment in patients with FA-HNSCC, although the same regimen proved toxic in patients with recurrent FA-HNSCC. The approval of EGFR TKIs for FA-HNSCC in Europe highlights the potential of these agents for this difficult to treat population.

Advances in the use of immune checkpoint inhibitors (ICIs) has dramatically changed the landscape of systemic recurrent/metastatic (R/M) HNSCC treatment. There are currently two FDA-approved ICIs for HNSCC, pembrolizumab and nivolumab, both of which were approved in 2016 (108). These mAbs target programmed cell death protein 1 (PD1) to stimulate lymphocyte-mediated antitumor activity, and can induce potent and durable responses in a subset of patients with HNSCC (109). ICI is now routinely used in combination with platinum and 5-FU, and in 2019, pembrolizumab was FDA approved as a first-line monotherapy for unresectable R/M HNSCC in patients with programmed death-ligand 1 (PD-L1)–expressing tumors (88). Patients who cannot receive first-line immunotherapy usually receive platinum-based combination chemotherapy incorporating cetuximab and a taxane or antifolate (90).

Because of the unique hematologic considerations of patients with Fanconi anemia, notably a high prevalence of allogeneic SCT, there is a general reticence to attempt cancer immunotherapy in this population. In particular, ICI administration post-SCT can trigger or exacerbate severe GVHD. A recent analysis of 150 patients without Fanconi anemia who received ICIs for hematologic malignancies after previous SCT found that 13% developed acute GVHD and 11% developed chronic GVHD, in addition to common hematologic side effects including neutropenia. However, of these patients with hematologic malignancy, an overall response rate of 48% was reported, with 28% achieving complete response and 20% demonstrating partial response (110). Although these responses may not recapitulate the response of patients with FA-HNSCC to ICI after SCT, it suggests that SCT itself does not universally portend a poor outcome to immunotherapy.

There are only three reports of patients with FA-HNSCC who received ICIs (all for disease recurrence; Table 7), of which one (Beckham patient No. 8) previously underwent SCT. Beckham patient No. 8 completed a course of palliative IMRT, after which he began nivolumab. However, after only three doses of 3 mg/kg, the patient was hospitalized for aspiration pneumonia and nivolumab-induced encephalitis, to which he succumbed (65). Beckham patient No. 9 received a single dose of 1 mg/kg tremelimumab (anti-CTLA4) and 20 mg/kg durvalumab (anti-PDL1) following three chemoradiation cycles with cetuximab and paclitaxel (described in the EGFR inhibitor section above). However, she died 10 days after administration of these ICI agents (65). The most recent report (Lewis 2020) detailed a case in which immune checkpoint inhibition with pembrolizumab was administered as part of adjuvant treatment for two locoregional recurrences of an oral cavity tumor. In this patient, pembrolizumab was administered every 3 weeks in combination with IMRT for the first recurrence and as a single agent for the second recurrence. ICIs were well tolerated in this patient, and was still being administered at the time of publication with no evidence of disease 4 months after the most recent surgery (97). Nonetheless, care should be taken in drawing conclusions based on so few reports, and there is a need for rigorous evaluation of the safety and efficacy of ICI in patients with FA-HNSCC. Furthermore, studies characterizing the prevalence and extent of PD-L1 expression in FA-HNSCC tumors are warranted. Only after substantial evidence of ICI toxicity profiles in this population is collected is it appropriate to conclude whether or not immune checkpoint inhibition is a viable treatment option for FA-HNSCC.

The paucity of aggregated reports or clinical trials describing the efficacy and toxicity of specific therapeutic regimens represents a major barrier in determining the appropriate treatment strategy for patients with FA-HNSCC. Complete surgical resection is accepted as the standard of care for these individuals given their heightened sensitivity to DNA-damaging agents including radiation and chemotherapy and the prevalence of oral cavity tumors which are commonly resected in sporadic HNSCC. However, surgery alone is rarely curative for advanced stage HNSCC and when the tumor arises in the hypopharynx or larynx, surgery may include total laryngopharyngectomy, which is dramatically life altering in young patients. We reviewed the case reports and case series of patients with FA-HNSCC reported in the peer-reviewed literature from 1966 through 2020 with a focus on descriptions of individuals who were treated with systemic chemotherapy and/or radiation in both adjuvant and definitive contexts for primary and/or recurrent disease. This review is intended as a guide for treatment discussions and decisions, and to stimulate the development of much needed clinical trials for this vulnerable population.

It is notable that the majority of patients with FA-HNSCC who received cetuximab experienced substantially less toxicity than those who underwent treatment with conventional cytotoxic chemotherapy. However, there are only nine total patients reported in the literature who received this EGFR-targeted mAb (five with primary tumors and four for recurrent disease), all of whom also received treatment with radiotherapy, ICIs, and/or cytotoxic chemotherapy agents (Beckham Patient Nos. 5, 7, 8, 9; Kutler patient Nos. 19, 21, 26; Nolan 2017; Wong 2013; refs. 4, 65, 86, 100). Of note, the majority of chemotherapy agents used were either cisplatin- or carboplatin-based regimens. There is an ongoing study assessing concurrent docetaxel and cetuximab with radiation in adjuvant treatment of HNSCC (RTOG 1216, NCT01810913), which may be better tolerated in patients with FA-HNSCC, although data in this population are lacking. One patient in this review received low-dose single-agent gemcitabine (100 mg/m² weekly) and another had low-dose 5-FU. Both patients tolerated the planned dose of radiotherapy, which suggests that alternative chemotherapy may need to be investigated further. Disentangling the toxicities of individual agents used in these regimens is difficult, and more information on the use of cetuximab in FA-HNSCC are needed to determine its role in disease management. However, given that the completion of radiotherapy is crucial for radiotherapy efficacy in HNSCC combined with the dismal tolerance of cytotoxic chemotherapy for FA-HNSCC, it seems prudent to consider cetuximab/radiotherapy for these patients.

Importantly, the one patient who received gefitinib experienced substantial regression of a stage IV oral cavity tumor without any notable toxicities reported during the treatment (Jung 2005; ref. 107). It is notable that she was the only FA-HNSCC patient who completed a course of systemic therapy alone, as all patients who received conventional chemotherapy without radiotherapy terminated treatment early due to toxicity. This case report is augmented by preclinical findings using small-molecule EGFR inhibitors in FA-HNSCC xenograft models in mice. In this study, gefitinib or afatinib significantly inhibited the growth of two distinct, cell line–derived xenograft models of FA-HNSCC, with in vivo toxicity studies in FANCA-deficient mice showing minimal side effects and no bone marrow toxicity (111). On the basis of these findings, the European Medicines Agency granted orphan drug designation to gefitinib (EU/3/18/2075) and afatinib (EU/3/18/2110) for the treatment of FA-HNSCC in late 2018. The remarkable evolution from data in one patient and a relevant preclinical model to drug approval is a testament to the power of single patient-studies for rare diseases. Additional evaluation of the efficacy of gefitinib and/or afatinib in patients with FA-HNSCC is critical to determine whether these drugs can improve survival and quality of life in this especially vulnerable population. The tolerability of cetuximab coupled with the approval of EGFR TKIs suggest that EGFR may represent a viable therapeutic target for FA-HNSCC. Studies to date in sporadic HNSCC have failed to identify predictive biomarkers for cetuximab, which was FDA approved in 2006 and is used in patients who cannot tolerate platinum chemotherapy. In addition, EGFR TKIs have proven ineffective in unselected HNSCC populations, although responses have been reported previously (112–114). Identification of the precise molecular alterations which drive HNSCC carcinogenesis in patients with Fanconi anemia remains crucial to developing effective targeted therapies with acceptable toxicity profiles.

ICIs are widely used in variety of cancers and recent studies suggest that defects in homology-dependent recombination could lead to higher predicted neoantigen load, increased tumor-infiltrating lymphocytes, and enhanced PD-1/PD-L1 expression (115). A recent report shows that patients with germline or somatic BRCA2 mutations are more likely to receive clinical benefit from ICIs across multiple tumor types (116). At least for patients with FA-HNSCC who have not received allogenic SCT, ICIs may be a reasonable option to consider, although more investigation is needed to elucidate PD-L1 expression along with the genomic landscape and tumor mutational burden in these patients. Studies are needed to determine whether prior SCT is an exclusion criterion for treatment with ICIs. Moreover, it is noteworthy that systemic agents are often reserved for recurrent or metastatic disease, which may be too late to observe benefit. Consideration of adjuvant therapy with curative intent immediately following surgery may reveal benefits that are inapparent in the face of widespread disease.

The collective evidence in these case reports and case series suggests that when possible, surgical resection with curative intent should remain the primary treatment modality for FA-HNSCC. Radiotherapy has been successfully administered with acceptable toxicity in the majority of cases in the modern era. There is likely no role for platinum-based cytotoxic chemotherapy. The role of non-platinum–based chemotherapy has not been explored and further investigation is needed. In lieu of platinum agents, EGFR inhibitors including cetuximab and TKIs may be both safe and effective. Immunotherapy may also be considered if the patient has not undergone SCT. The suggested FA-HNSCC treatment pathway is summarized in Fig. 2.

Finally, it is crucial to reiterate the importance of physician awareness in detecting and treating FA-HNSCC. The variability (or absence) of physical anomalies in patients with Fanconi anemia requires a high degree of suspicion for all teenagers or young patients presenting with head and neck cancers without established risk factors. Conversely, known Fanconi anemia carriers should undergo frequent dedicated screening examinations for cancer. The recognition of FA-HNSCC as a distinct entity will further allow individual institutions to compile and share treatment strategies and outcomes with the entire medical community. With global collaboration, we can begin to construct an accessible database complete with safety and efficacy data, which would provide an invaluable resource for guiding thoughtful, evidence-based therapies for future patients with Fanconi anemia with HNSCC.

H. Kang reports personal fees from Bayer, Achilles Therapeutics, MitoImmune, PIN Therapeutics, GSK, Prelude Therapeutics, and Genentech; and grants from Elevar Therapeutics, Kura Oncology, Exelixis, Eli Lilly, NeoImmune Tech, and PDS Biotechnology outside the submitted work. S.S. Yom reports grants from Bristol Myers Squibb, Merck, Genentech, and BioMimetix outside the submitted work. A. Smogorzewska reports grants from Rocket Pharmaceuticals outside the submitted work. D.E. Johnson reports grants from NIH R01 DE023685, NIH R35 CA231998, and NIH R01 DE028289 during the conduct of the study; and nonfinancial support from STAT3 Therapeutics outside the submitted work. J.R. Grandis reports nonfinancial support from STAT3 Therapeutics outside the submitted work. No disclosures were reported by the other authors.

This work was supported by NIH grants R01 DE023685 (J.R. Grandis and D.E. Johnson), R35 CA231998 (J.R. Grandis), R01 DE028289 (D.E. Johnson and J.R. Grandis), R01 CA204127 (A. Smogorzewska), and V Foundation translational grant T2019-013 (A. Smogorzewska). A. Smogorzewska is an HHMI Faculty Scholar.