INTRODUCTION

Germline pathogenic/likely pathogenic (P/LP) variants in PALB2 (Partner and Localizer of BRCA2) were first associated with increased cancer risk in 20071,2,3 and clinical testing has been available since then. Testing for PALB2 increased with its inclusion on multigene cancer panels starting around 2012–2013. It has come to be considered as the third most important breast cancer gene after BRCA1 and BRCA2 following the 2014 publication of robust breast cancer risk estimates that overlap with BRCA2.4 Despite the emerging importance of this gene, there is a dearth of guidelines regarding the clinical management of women and men with PALB2 germline P/LP variants (henceforth called “heterozygotes”). Given the recently published updates on PALB2-associated cancer risks,5,6 there remains a gap in implementing this information to optimize patient care. Consequently, it is a timely opportunity to translate these and other findings into clinical practice for PALB2 heterozygotes.

The recent international effort to estimate cancer risk for PALB2 heterozygotes was based on 524 families,5 and highlighted the need to develop a practice resource through the lens of an international workgroup, through synthesis and practical application of existing data to guide clinical practice. BRCA1 and BRCA2 have long been considered “high risk” breast cancer genes, while ATM and CHEK2 have been considered “moderate risk” breast cancer genes. However, PALB2 clearly blurs these distinctions given that the range of breast cancer risks associated with PALB2 overlaps with risks associated with “high” and “moderate” risk genes.7 Through this Clinical Practice Resource, we provide guidance on personalized risk estimation, especially through the use of CanRisk-BOADICEA,8,9 the only risk estimation tool that currently incorporates PALB2.

MATERIALS AND METHODS

After proposal approval and review of potential or actual conflicts as per the relevant American College of Medical Genetics and Genomics (ACMG) policies, the workgroup developed a list of clinical areas in the management of PALB2 heterozygotes. Workgroup members were identified with expertise in clinical cancer genetics, breast and gynecologic surgery, and medical oncology from Australia, Asia, the United States, Canada, the United Kingdom, and Europe. After approval of a proposal by the ACMG Board, workgroup members assembled relevant peer-reviewed publications based on knowledge of the existing literature and performed additional PubMed searches (as of January 2021). Publications that investigated populations, large clinic cohorts (more than 100 heterozygotes, preferably from multiple institutions), and large case series were prioritized; however, we acknowledge that not using a systematic evidence review approach, while pragmatic, may have led to some pertinent literature being missed. Clinical management recommendations were derived by consensus from this literature (through monthly teleconference calls and email review) and the collective expertise of the authors. Working and final drafts were reviewed and approved by members of the Professional Practice and Guidelines Committee and the ACMG Board of Directors.

Risk estimation

The initial studies in 2007 that identified an increased risk of breast cancer in PALB2 heterozygotes were subsequently confirmed by multiple small studies, culminating in an international collaboration of 154 families by the PALB2 Interest Group,4 which has recently been expanded to 524 families.5 The estimated risk of female breast cancer to age 80 years was 53% (95% confidence interval [CI]: 44–63%). The relative risk for a female PALB2 heterozygote born between 1950 and 1959 was 7.18 (95% CI: 5.82–8.85); this relative risk increases in women from more recent birth cohorts. The same study reported a modest increased risk of ovarian cancer of 0.6% (95% CI: 0.3–1.3%) to age 50 and 4.8% (95% CI: 2.4–9.7%) to age 80, with findings confirmed in a case–control study (n = 14,135 cases) that estimated the risk to age 80 to be 3.2% (95% CI: 1.8–5.7%).6 Two recent, very large breast case–control studies estimated the odds ratio (OR) for PALB2-related breast cancer risk to be 3.83 (95% CI: 2.68 to 5.6310 and 5.02 (95% CI: 3.73–6.76),11 with no increased risk for missense variants.11 The absolute risks of developing breast or ovarian cancer are predicted to be influenced by cancer family history.5 For example, the estimated absolute risk of developing breast cancer by age 80 years varies from 52% (95% CI: 42–62%) for a female with an unaffected mother at age 50 years and unaffected maternal grandmother at age 70 years to 76% (95% CI: 69–83%) for a female with two affected first-degree relatives.5 Similarly, the estimated risk of developing ovarian cancer by age 80 years varies from 5% (95% CI: 2–10%) for a female with no family history of ovarian in first and second-degree relatives to 16% (95% CI: 8–28%) for a female whose mother and sister developed ovarian cancer at age 50 years. These risk estimates are also modified by polygenic modifiers (see below) that can be combined with lifestyle factors to give a personalized risk estimate such as CanRisk.8 There are currently no prospective risk estimates for contralateral breast cancer; one study estimated the 5-year cumulative incidence of contralateral breast cancer to be 10%, but this was retrospective and based on small numbers.12 The risks for pancreatic cancer in PALB2 heterozygotes are estimated to be 2–3% (95% CI: females: 1–4%; males: 2–5%) to age 80 years compared with 1.5% in the general population. For male breast cancer the risk is 1% (95% CI: 0.2–5%)5 compared with a general population risk of less than 0.1%. Germline PALB2 P/LP variants have been detected in a few gastric cancer cases,13,14 but it is not known if this is a true association. There is no consistent evidence for increased risks in prostate or colorectal cancer. Although germline PALB2 P/LP variants have been reported in children with certain cancers (see below), no large-scale studies providing estimates of risk are available.

Polygenic risk score and other modifiers

Variation elsewhere in the genome may modify cancer risks associated with PALB2 P/LP variants. Genome-wide association studies have identified large numbers of common genomic variants that contribute to an individual’s risk of breast cancer. The cumulative effect of these individually minor risks is summarized in a polygenic risk score (PRS) and numerous studies have shown that this measure provides effective stratification of breast cancer risk in the general population.15 For some breast cancer predisposition genes, studies have demonstrated that the risk associated with a P/LP variant combines with the risk from the PRS in a multiplicative fashion,16,17 and this may be a general principle, although evidence to support this assertion is needed. For single genes with more moderate effect, modification by the PRS is sufficient to change an individual’s final risk classification and their corresponding clinical management. Information on polygenic risk has come predominantly from studies restricted to populations of European ancestry and the extent to which this can be applied for women with other ethnic backgrounds is unclear,18 although recent studies have found that the established PRS has some value at least in Asian populations.19 Few studies have directly examined PRS modification in women with P/LP variants in PALB2, and the most significant study to date examined the effect of a PRS incorporating 86 common variants in a large cohort undergoing testing for multiple cancer associated genes that included 906 heterozygotes of PALB2 pathogenic variants.7 In this group, the strength of the association of the PRS with breast cancer risk was the same (or only slightly reduced) in women who harbored PALB2 P/LP variants as measured in nonheterozygotes. When factoring in PRS, mammographic density, and lifestyle/hormonal risk factors, 8% of heterozygotes would have a lifetime risk less than 30%, 63% of heterozygotes would fall in the 30–60% range, and 29% of heterozygotes would have BRCA1/BRCA2 equivalent lifetime risks of greater than 60%.8 The modifying effect of the PRS is essentially independent of the risk associated with family history and these and other conventional risk factors can be combined into a single personalized risk assessment. When and if PRS becomes available in the clinical setting this integrated approach can be implemented in some online assessment tools such as CanRisk,8,20 although there are no data yet that directly demonstrate improved clinical outcomes from the greater individualization of risk.

  • ACMG recommends the use of personalized risk estimates (e.g., CanRisk) in guiding clinical management.

Indications for genetic testing

Testing for germline P/LP PALB2 variants is usually done as part of a wider gene panel, e.g., for breast, ovarian, and/or pancreatic cancer. Case studies illustrating PALB2 testing in practice are shown in Fig. 1. Overall indications for genetic testing for inherited breast cancer per these guidelines are based on personal and/or family cancer history, taking into account age at diagnosis (at or below age 45), triple-negative breast cancer (at or below age 60), or presence of another primary cancer or family history of cancers. A diagnosis of ovarian or pancreatic cancer in an individual is sufficient to meet current National Comprehensive Cancer Network (NCCN) guidelines for germline testing.21 In addition, germline testing should be offered when a PALB2 variant is identified through tumor testing. Genetic and oncology professional societies have endorsed consenting for the return of P/LP germline variants detected as part of tumor sequencing (American Society of Clinical Oncology [ASCO];22 ACMG23). For PALB2, this is especially relevant given the increasing availability of promising targeted therapies contingent on a P/LP germline variant, such as poly(ADP)-ribose polymerase (PARP) inhibitors.

  • ACMG recommends that PALB2 should be included in breast, ovarian, and pancreas germline cancer gene panels.

Fig. 1: Case studies illustrating PALB2 testing in practice.
figure 1

(a) A 59-year-old female without a prior history of cancer had cascade testing through her gynecologist, which identified a pathogenic variant in PALB2, c.695del (p.Gly232fs), originally identified in her 90-year-old mother who was diagnosed with colon cancer at age 88 and had a prior diagnosis of breast cancer at age 64. The patient was advised by her gynecologist to have a risk-reducing oophorectomy and presented for a second opinion to a hereditary cancer clinic for guidance. She was having annual mammograms and breast magnetic resonance images (MRIs) with normal results, as well as colonoscopy every five years, due to the family history of colon cancer, with two benign colon polyps detected during the last one. Discussion points: (1) In the absence of a family history of ovarian cancer, a risk-reducing salpingo-oophorectomy is not typically recommended in women with a PALB2 P/LP variant who would have a lifetime risk of ovarian cancer of 3–5%. (2) Management of raised colorectal cancer risk is based on the family history, not PALB2 carrier status. (3) The patient does not meet guidelines for pancreatic cancer surveillance, given lack of family history for this malignancy. (b) A 70-year-old female referred to clinic for genetic evaluation and testing, following diagnosis of metastatic pancreatic cancer treated with chemotherapy, with a strong family history of breast cancer. Tumor testing identified one PALB2 pathogenic variant c.3113G>A (p.Trp1038Ter), which was confirmed following genetic evaluation and subsequent germline testing. Discussion points: (1) If the son has inherited the pathogenic variant, pancreatic cancer surveillance could be considered, as he would meet both International Cancer of the Pancreas Screening (CAPS) and National Comprehensive Cancer Network (NCCN) guidelines for individuals in whom surveillance is considered reasonable. Ideally, such surveillance is done in the context of a research study. (1) If any of the maternal cousins have inherited the pathogenic variant, current NCCN guidelines do not recommend pancreatic cancer surveillance for PALB2 heterozygotes and 3rd degree relatives with pancreatic cancer. (2) The family history of breast cancer puts this family at the higher range of risk estimates, based on modifying risks. (3) The identification of PALB2 may guide treatment options including use of platinum-based treatment or poly (ADP-ribose) polymerase (PARP) inhibitors, some of which may be through clinical trials.

Full size image

Variants of uncertain significance

The identification of variants of uncertain significance (VUS) in PALB2 represents a considerable clinical challenge. Published studies on PALB2 penetrance and risk have focused on predicted protein truncating variants (frameshift, nonsense, splice, exonic deletions/duplications), and the first PALB2 missense variant to be determined as pathogenic was c.104 T>C (p.Leu35Pro), which abrogates the PALB2–BRCA1 interaction and disables its abilities to promote homologous recombination.24 Three groups have recently published their experiences with the use of a range of functional assays to characterize PALB2 missense variants25,26,27 (summarized by Southey and colleagues28). The combined functional and epidemiological evidence published to date suggest that only a small minority of known missense variants are potentially pathogenic. The ClinGen Hereditary Breast, Ovarian and Pancreatic Cancer Variant Curation Expert Panel (https://clinicalgenome.org/affiliation/50039/) is working to define specifications of the ACMG and the Association for Molecular Pathology (AMP) rules for missense variants including integration of functional data. A VUS should not be used to guide clinical management but must be periodically re-reviewed to determine changes in interpretation based on emerging data.

  • ACMG recommends that PALB2 VUS are not used to guide clinical management.

Pathology and outcomes

The initial reports of the pathological features of PALB2-related breast cancers were based on the Finnish founder variant that accounts for ~0.7% of all breast cancer in Finland. This variant, c.1592delT p.(Leu531fs), was strongly associated with high-grade triple-negative breast cancer (TNBC) (55% vs. 9.4% for nonfamilial breast cancer).29 Subsequent studies, while reporting that PALB2-related breast cancers were usually high grade, did not find as high a frequency of TNBC,12,30,31,32,33 but recent and very large commercial sequencing studies have shown that TNBC cases are substantially enriched (1.4%) for germline pathogenic variants in PALB2.34,35 This relationship also holds true in Asian breast cancer patients31,36 and in African American breast cancer patients, where there is a particularly strong association with TNBC (OR 23.5, P < 0.001) (by comparison, the OR for the TNBC association with BRCA1 pathogenic variants in this population was 180).37

Recent molecular studies have shown that as for BRCA1 and BRCA2, biallelic inactivating pathogenic variants in PALB2 (seen in two-thirds of breast cancer tumors with germline pathogenic variants in PALB2) are nearly always associated with mutational signatures associated with defects in homologous recombination repair deficiency (HRD).32,33,38,39,40 Li et al. identified patterns and frequencies of somatic alteration in PALB2-related breast cancer that distinguished them from The Cancer Genome Atlas (TCGA) breast cancer cases of the same immunohistochemical phenotype.32 Overall, PALB2-related breast cancers appear to be more biologically similar to BRCA2- than to BRCA1-related breast cancer,3,32,33 and all three forms of hereditary breast cancer are much more alike to each other than any of them are to non-HRD-related familial breast cancer, or to breast cancer in general.

How do these pathology features influence outcome following a breast cancer diagnosis? In the first study of outcome, the survival following breast cancer appeared to be significantly worse for familial PALB2 cases than for either sporadic or familial (non-BRCA1, BRCA2, or PALB2) breast cancer.29 In support of this, a study of metastatic compared with early breast cancer noted that PALB2 was one of eight genes more frequently mutated in metastatic than early breast cancer.41 A recent study of nearly 3,000 unselected breast cancers in China showed that nearly 1% carry a PALB2 pathogenic variant; overall survival of these women was significantly worse compared to those with no PALB2 pathogenic variant.31 The most comprehensive analysis of survival was based on two founder pathogenic variants in Poland, where 12,529 women with breast cancer were genotyped. Nearly 1% were found to have one of these two variants, compared with 0.2% in controls. The 10-year survival for PALB2 heterozygotes was only 48%, compared with 75% for those without these variants (hazard ratio [HR] for death = 2.27, P < 0.0001).12 Despite these troubling findings, HRD-high tumors may respond better to chemotherapy,33 so it will be important to include variant status of PALB2 to interpret results of some targeted therapies in breast cancer clinical trials.

  • ACMG recommends prospective collection of clinical data from PALB2 heterozygotes to establish clear metrics on treatment outcome and survival.

Surveillance and risk-reducing surgery

Breast cancer risks have now been well documented in female PALB2 heterozygotes, thus establishing PALB2 as a major cancer susceptibility gene. Therefore, recent guidelines such as the NCCN guidelines for genetic familial high-risk assessment,21 Australian national oncology guidelines (eviQ)42 the European Society of Medical Oncology,43 and the German S3 guideline for breast cancer44 consider PALB2 a moderate-to-high-risk gene for breast cancer. Although the clinical utility of preventive measures as outlined in the analytical validity, clinical validity, clinical utility, and ethical, legal and social implications (ELSI) (ACCE) model is not sufficiently proven yet,45 there is a demand for clinical intervention that needs to be addressed. Accordingly, the NCCN guidelines include recommendations for breast cancer surveillance for female PALB2 heterozygotes. Also, ASCO, the American Society for Radiation Oncology, and the Society for Surgical Oncology Guideline have recently published breast management guidelines for patients who carry P/LP variants in hereditary breast cancer genes.46 Based on published risks, intensified breast cancer surveillance for women carrying P/LP PALB2 variants is recommended. This includes early onset of surveillance including mammograms, tomosynthesis, and magnetic resonance images (MRIs) with contrast starting at the age of 30 years. The optimal surveillance strategy remains to be determined with regard to the addition and frequency of mammograms and age of termination of surveillance. Importantly, clinical utility of intensified surveillance with regard to key surrogate markers, e.g., positive and negative predictive values of the surveillance strategy and hard endpoints, i.e., mortality and morbidity reduction, is largely missing and may depend on the specific phenotype of PALB2-associated breast cancer.

Among PALB2 heterozygotes, NCCN recommends discussion of risk-reducing bilateral mastectomy, with nipple-sparing mastectomy as an option (www.asco.org/breast-cancer-guidelines).21 Comprehensive counseling including a three-generation pedigree, tailored cancer risk assessment, the personal life situation and preferences of the counselee; all need to be considered when counseling for prophylactic surgery. Nongenetic risk factors such as dense breast tissue and hormonal/lifestyle modifiers may be included.8,15 Importantly, age-specific risks (e.g., within the next ten years) should be communicated to allow counselees to make a decision over a manageable period of time. Competing risks due to other health risks and pre-existing conditions need to be considered. Given the lack of data on contralateral breast cancer risk, with regard to contralateral mastectomy, the person’s current circumstances and competing risk factors should be considered, and a shared decision-making approach should be employed. Breast cancer surveillance in those with a previous breast cancer who have remaining breast tissue should be performed according to the guidelines for healthy heterozygotes.

Establishing recommendations for PALB2-associated ovarian cancer risks has been more challenging. Two recent studies have estimated risk to 80 years to be 3.2% (95% CI: 1.8–5.7%)6 and 5% (95% CI: 2–10%)5 compared with a population-based lifetime risk of 1.5–2%. Both studies estimated the risk to age 50 years to be well under 1%. Currently, the NCCN guidelines acknowledge that PALB2 heterozygotes have an increased risk for ovarian cancer and cite insufficient evidence, based on variant status alone, to recommend risk-reducing salpingo-oophorectomy (RRSO).21 Based on this information, risk-reducing bilateral salpingo-oophorectomy should be considered in a nondirective counseling process taking additional risk and protective factors into consideration as outlined above. In case of a decision in favor of RRSO, performing the procedure at or after menopause may be appropriate, considering that the risk before this is very small. For patients undergoing tubal ligation for contraception, opportunistic salpingectomy could be considered.47 The benefit of ovarian cancer surveillance through the use of pelvic ultrasound and CA-125 levels is considered insufficiently sensitive for early detection of ovarian cancer and therefore not recommended. Women should be counseled regarding the limitations of this surveillance.

The NCCN and the International Cancer of the Pancreas Surveillance (CAPS) Consortium48 recently updated pancreatic cancer surveillance recommendations, through annual MRI or magnetic resonance cholangiopancreatography (MRCP) and/or endoscopic ultrasound (EUS). Specifically, per NCCN guidelines, if a first- or second-degree relative is affected with pancreatic cancer, surveillance may be considered beginning at age 50.21 Per CAPS recommendations, if a first-degree relative is affected with pancreatic cancer, surveillance may be considered beginning at age 45–50.48 It is important to recognize that pancreatic cancer surveillance recommendations are mainly based on consensus rather than more rigorous evidence assessments, as additional data continue to be collected to determine benefits from surveillance. Although in the United States, surveillance for pancreatic cancer is encouraged in the context of a surveillance study, the position in the United Kingdom is based upon a lack of current data to support efficacy, thus pancreatic cancer surveillance is not recommended outside of a research study.

ACMG recommends

  • Surveillance for breast cancer should be equivalent to that for BRCA1/2 heterozygotes.

  • Risk-reducing mastectomy can be considered as an option. The decision should be guided by personalized risk assessment.

  • Ovarian cancer surveillance should not be offered, and risk-reducing salpingo-oophorectomy should include shared decision making and should rarely be considered before the age of 50.

  • Pancreatic cancer surveillance should be considered, but ideally as part of a clinical trial.

Therapeutic implications of PALB2 gene variation

Inactivating variants in PALB2, when associated with biallelic inactivation in the tumor (i.e., loss of heterozygosity or biallelic variants) confer a deficiency in the homologous recombination pathway.32,49 This molecular phenotype makes these tumors more vulnerable to DNA-damaging agents, including platinum-based chemotherapy. In this regard, several case reports have described remarkable clinical activity of platinum-based chemotherapy in patients with PALB2-associated advanced breast or pancreatic cancer.50,51,52,53 Some of the prior studies have retrospectively demonstrated an extended progression-free survival and even overall survival when incorporating platinum-based chemotherapy in patients with pancreatic ductal carcinomas harboring BRCA1/2 or PALB2 variants.

Similarly, cells with HRD are exquisitely sensitive to poly(ADP)-ribose polymerases (PARP), which prompted the clinical development of PARP inhibitors (PARPi) in patients with BRCA1 and BRCA2 variants and later to a wider group of patients harboring dysfunctional homologous recombination repair. A preclinical in vivo study performed in patient-derived xenografts (PDX) from patients with breast cancer showed that PDXs from PALB2 germline variant heterozygotes were homologous recombination repair pathway deficient as assessed by nuclear RAD51 foci. Additionally, all 11 breast cancer samples from patients carrying a germline PALB2 variant scored RAD51 foci negative, confirming this deficiency and providing evidence to clinically develop PARPi in this group.49

To date there are no randomized controlled trials specifically targeting patients with pathogenic germline PALB2 variants. Given the mechanistic similarities and presentation with BRCA1/BRCA2-associated cancers, it is not surprising that oncologists have in recent years used similar regimens as those that have been successful in patients with germline BRCA1/2 P/LP variants (see Supplementary Table 1). Notably, some pancreatic adenocarcinoma trials have included patients with pathogenic germline PALB2 variants alongside BRCA1/2 patients, and some trials in patients with metastatic breast cancer and an associated germline variant or somatic variant in homologous recombination repair pathway genes beyond germline BRCA1/2 have also provided very promising data among PALB2 heterozygotes.54,55 Within the talazoparib trial, 13 patients had breast cancer and five had a germline PALB2 variant.54 Among them, four had high HRD scores by Myriad MyChoice genomic score and three had a partial response after treatment with talazoparib. In the olaparib trial, of 11 patients with advanced breast cancer and a PALB2 variant, 9 achieved a partial response (82%) and 2 had stable disease.55 Overall, the median duration of response was nine months. Interestingly, the majority were ER+/HER2- and one was HER2+. These preliminary findings provide strong rationale to warrant further clinical development of PARPi in this population. Finally, among patients with advanced prostate cancer and homologous recombination repair pathway deficiency, clinical evidence is accumulating to show that PARPi are effective in this population where PALB2 variants achieved the second highest response rate to olaparib behind BRCA1/2 variants.56,57

  • ACMG recommends PALB2 heterozygotes should be considered for the same therapeutic regimens and trials as those for BRCA1/2.

Genetic counseling

Despite the increasing recognition of the importance of PALB2 predisposing to inherited breast cancer, the evidence to support cancer risks and management remains understandably less developed compared with that for BRCA1 and BRCA2. As a result, there remain challenges in genetic counseling, risk assessment, and sharing management recommendations, as clinicians outline both what is known and what is not yet known to these patients, to guide them to make the best decision for themselves. Additional complexities include the range of risks that cross the threshold between moderate and high-penetrance genes. Consequently, PALB2 may be considered a prototypic gene to highlight the significant limitations in categorizing genes according to high versus moderate penetrance, and it provides an ideal lens through which to develop consensus and a framework for how to think about risks as a continuous rather than categorical (or discrete) variable. It follows that these factors also impact the use of established risk reduction strategies (e.g., risk-reducing mastectomy), originally implemented for genes categorized as high penetrance (e.g., BRCA1/2), compared with surveillance, which is generally the risk management strategy recommended for genes categorized as moderate penetrance (e.g., ATM, CHEK2). PALB2 heterozygotes or those with a family history of such are advised to follow up with a genetics provider periodically for updates. There is no established genotype–phenotype correlation. PALB2 has recently been added to the ACMG Secondary Findings v3.0 list.58,59

Fanconi anemia

Biallelic germline L/LP variants in the homozygous or compound heterozygous state in PALB2 is a very rare cause of Fanconi anemia, FA-N (incidence less than 1 in 3 million), a childhood-onset condition associated with bone marrow failure, physical abnormalities, organ defects, and an increased risk of certain cancers.60,61 The high risk for certain childhood cancers in FA-N was soon recognized including medulloblastoma60,62 and Wilms tumor.60,63 Large-scale studies have uncovered P/LP PALB2 heterozygous variants in osteosarcoma, leukemia, brain tumors and soft-tissue sarcoma,64 and pediatric high-grade glioma.65 The detection of germline PALB2 P/LP variant(s) in a child with Fanconi anemia and/or childhood cancer offers an opportunity for cascade testing to determine whether other adult family members are at risk. Conversely, while genetic counseling of individuals with a P/LP PALB2 variant may include discussion of biallelic inheritance (and implications to family planning with consideration of partner testing prior to achieving pregnancy), in reality, outside of countries with a founder variant, this risk is very small (assuming a PALB2 carrier frequency of 1 in 700, the probability of a liveborn offspring with Fanconi anemia will be less than 1 in 2,800 as some P/LP variant combinations are likely to be embryonic lethal). Additional family planning considerations include discussion of preimplantation genetic diagnosis to identify P/LP PALB2 variants at the embryo stage. Generally, prenatal diagnosis through amniocentesis or chorionic villus sampling is not typically offered for detection of a P/LP PALB2 variant in the fetus, given that this is an adult onset cancer predisposition syndrome.

  • ACMG does not recommend testing partners of PALB2 heterozygotes in the reproductive setting, unless they are from a country with founder variants or it can be justified by the partner’s family history of cancer.

Research gaps in clinical areas of need

There remains a paucity of data on PALB2 heterozygotes compared with BRCA1/BRCA2, both in terms of cancer incidence, spectrum, and clinical outcomes. This extends to establishing the full spectrum of childhood cancer risk in PALB2-related Fanconi anemia. Genome-first and/or population-scale sequencing approaches could be used to improve risk estimates and the clinical application of polygenic scores will further refine risk estimates. Prospective data collection is needed to determine the efficacy of surveillance and risk-reducing surgery as well as establishing the contralateral breast cancer risk. Efforts should be addressed at better histopathological and molecular characterization of PALB2-related cancers, and how this influences clinical outcome. The role of chemoprevention needs to be established, and PALB2 heterozygotes should be eligible for the same or equivalent therapeutic studies as BRCA1/BRCA2. PALB2 heterozygotes are much less common than BRCA1/2 outside of countries with founder populations, but large international collaborations make it feasible to collect enough data to facilitate evidence-based management approaches.

Conclusion

The recommendations made here have been based on expert opinion using comprehensive literature ascertainment approach, but not systematic review. There is strong evidence that P/LP PALB2 variants confer a range of breast cancer risks across what is considered moderate to high; consequently, enhanced surveillance and the option of risk-reducing interventions are warranted. The risk range for this gene underlies the need to move away from compartmentalizing PALB2 and consider risk to be a continuous variable from high to moderate, influenced by family history, polygenic risk score, and other factors.7 The same applies to other breast cancer genes. Changing this paradigm will allow us to move to personalized risk estimates by placing the risk from the P/LP variant in the context of other risk factors and develop strategies to translate this information to enhance medical management. There is reasonable evidence that PALB2 P/LP variants confer a small to moderately increased risk for ovarian cancer that may warrant risk-reducing interventions, albeit their clinical benefit is not sufficiently proven yet with respect to the efficacy of preventive measures to reduce morbidity and mortality. Likewise, there is reasonable evidence that such variants confer a small to moderately increased risk of pancreatic cancer, but the role of surveillance remains controversial. Given the many uncertainties, those at risk for PALB2-related cancers, and the health professionals who care for them are encouraged to contribute follow-up data to long term studies, thereby facilitating the generation of prospective cancer risk estimates and the evaluation of prevention measures. Current evidence supports the consideration of platinum-based regimens and clinical trials of PARPi in patients with germline P/LP PALB2 variants and breast, ovarian, prostate, or pancreatic cancer, especially when biallelic inactivation and HRD are present.