Response to bipolar androgen therapy and PD-1 inhibition in a patient with metastatic castration-resistant prostate cancer and a germline CHEK2 mutation

Abstract

We present the case of a patient with germline CHEK2-mutated metastatic castration-resistant prostate cancer (mCRPC) who responded to bipolar androgen therapy (BAT) combined with pembrolizumab after progressing through multiple lines of therapy. The patient was diagnosed in his 40s following an elevated screening prostate-specific antigen and biopsy. Over the course of 20 years, he progressed through nearly all standard therapies including androgen deprivation, combined androgen blockade, traditional chemotherapy, targeted therapies and experimental agents. He was ultimately treated with BAT, whereby the patient’s cycle was between low (castrate) and high (supraphysiological) testosterone levels. This counterintuitive approach resulted in a marked response to BAT plus pembrolizumab consolidation lasting 13 months. His underlying germline mutation in CHEK2, an important mediator of DNA repair, may have sensitised the cancer cells to the DNA damage caused by BAT. Single case report outcomes should not be used as evidence of efficacy for treatment regimes. Our case supports further investigation into BAT plus immunotherapy for patients with DNA repair-deficient mCRPC.

Background

Prostate cancer is the second most common cancer diagnosed in men, with an estimated global incidence of 1 400 000 cases per year leading to 723 000 deaths annually.¹ Androgen deprivation therapy (ADT) is the backbone of treatment of prostate cancer, though most patients will progress through this modality and require additional therapeutic agents. When metastatic progression develops despite castrate levels of testosterone, this state is described as metastatic castration-resistant prostate cancer (mCRPC), which is typically treated with novel anti-androgen therapy in the frontline, or taxane chemotherapy for men previously exposed to ADT and androgen receptor (AR) inhibitors in earlier hormone-sensitive disease states. Resistance to ADT and additional anti-androgen therapies is thought to be mediated primarily by adaptive upregulation of AR activity, either through genomic amplification, AR enhancer gain, AR mutations and AR splice variants leading to constitutive signalling.² Bipolar androgen therapy (BAT), whereby patients are exposed to rapid cycles of castration levels of testosterone alternating with supraphysiological levels of testosterone, has emerged as a strategy to overcome AR therapy resistance, although the mechanism of BAT efficacy remains unclear.³ Immunotherapy in the form of PD-1 checkpoint blockade has also been used in mCRPC,^{4 5} though responses are more limited to subsets of patients with mCRPC and mismatch repair deficiencies.^{6 7} Approximately 12%–20% of patients with metastatic prostate cancer have germline or somatic mutations in DNA repair genes.⁸ CHEK2 is one such gene, which codes for a checkpoint kinase important in homologous DNA repair, that also has a unique negative regulatory relationship with the AR.⁹ Prior studies have shown only limited clinical efficacy of the PARP inhibitor olaparib in CHEK2-mutated patients with mCRPC,¹⁰ and thus these patients have major unmet medical needs. We describe a patient with mCRPC and a germline CHEK2 mutation, who had progressed through several lines of therapy, and demonstrated a response to BAT and immunotherapy with pembrolizumab.

Case presentation

A man in his 40s undergoing routine health screening was found to have a prostate-specific antigen (PSA) level of 10.4 ng/mL (figure 1). Prostate biopsy revealed Gleason grade 7 (4+3) adenocarcinoma. He received definitive local therapy with Iodine-131 brachytherapy and external beam radiotherapy. PSA levels nadired 3 years later at 0.2 ng/mL, but then he developed biochemical recurrence 5 years later when his PSA gradually rose to 2.7 ng/mL. Multimodal imaging at that time was unremarkable and repeat prostate biopsies were negative. The PSA continued to rise over the next 2 years to 7.8 ng/mL, and ADT with leuprolide was initiated. PSA decreased to 0.1 ng/mL, but CT imaging showed a new soft tissue mass at the left distal ureter, which was biopsied and found to be a lymph node harbouring metastatic prostate cancer. The patient’s PSA rose to 0.40 mg/dL so bicalutamide was added for his mCRPC and he was treated with sipuleucel-T.

Figure 1.

Prostate-specific antigen over time. (1) Maintenance with androgen deprivation therapy (leuprolide). (2) Combined androgen blockade with addition of bicalutamide. (3) Surgical resection (left nephroureterectomy, partial cystectomy, lymph node dissection). (4) Resumed combined androgen blockade with leuprolide and bicalutamide. (5) Stopped bicalutamide, started abiraterone (with continued leuprolide). (6) Started olaparib. (7) Changed to cabazitaxel. (8) Initiated bipolar androgen therapy. (9) Started stereotactic body radiation therapy to metastatic disease. (10) Discontinued bipolar androgen therapy and started enzalutamide. (11) Admitted with pain and dyspnoea, found to have rapid progression in metastatic soft tissue sites, transitioned to hospice. Figure created by Benjamin T Berger.

After an initial PSA response, the PSA rose again and interval imaging showed left-sided hydronephrosis with obstruction at the ureterovesical junction and thickening of the bladder wall. The patient underwent left nephroureterectomy, partial cystectomy, partial left seminal vesiculectomy and extended pelvic lymph node dissection. Pathology showed prostatic adenocarcinoma of the left seminal vesical invading the left distal ureter and bladder, without lymph node involvement. PSA nadired at 0.01 ng/mL and the patient was observed off ADT for the next 2 years until the PSA rose to 29.23 ng/mL, with a corresponding testosterone level of 284 ng/dL (from 39 ng/dL the year prior). Imaging showed enlarging inguinal and retroperitoneal nodes consistent with M1a mCRPC development. Leuprolide was continued and abiraterone acetate with low-dose prednisone were started.

Tumour genetic sequencing (Foundation One, Cambridge, Massachusetts, USA) was performed and was notable for a CHEK2 T367fs* frameshift mutation, a MUTYH E396fs*43 frameshift mutation and a TP53 R158H missense mutation. The tumour was determined to be microsatellite stable with a low mutation burden. Germline testing for hereditary prostate cancer genes was also performed (GeneDx) and was notable for a heterozygous pathogenic variant in CHEK2 (p.T367fs*), which is known to be an Eastern European founder mutation.¹¹

The patient’s PSA decreased to 6.1 ng/mL, but his disease again progressed radiographically. He was enrolled in a clinical trial of the EZH2 inhibitor lirametostat (CPI-1205) combined with enzalutamide until progression and he did not respond. He subsequently received four cycles of docetaxel chemotherapy until he had PSA progression and enlargement of his retroperitoneal lymph node. He was transitioned to olaparib, and his PSA declined from 51.39 ng/dL to 35.18 ng/dL in 2 weeks. Progression was observed 8 weeks later, however, with the PSA rising to 64.47 ng/dL. Cabazitaxel chemotherapy therapy was initiated. Upon further progression, he received carboplatin as well as palliative radiation therapy. His PSA declined to 49.42 ng/dL, but CT of the abdomen and pelvis showed multiple worsening sites of disease with overall increasing lymphadenopathy in the supraclavicular, para-aortic and retroperitoneal spaces, and a new right adrenal mass. Hospice care was being considered. The patient sought experimental or off-label therapies that might delay or prevent his symptomatic deterioration. After informed decision-making, he elected to pursue off-label BAT plus pembrolizumab based on phase 2 clinical studies demonstrating benefits in selected patients.^{3 12}

The patient started BAT, with testosterone cypionate 400 mg intramuscularly every 28 days, while continuing ADT with leuprolide 45 mg intramuscularly every 6 months. After 3 months of therapy, his PSA had decreased to 1.55 ng/mL and CT imaging showed a significant interval decrease in the size of the adrenal mass and the supraclavicular, mediastinal and retroperitoneal lymphadenopathy (figure 2). Consolidation therapy with pembrolizumab in addition to BAT, as per the COMBAT trial¹² but off-label in this patient, was then pursued. Due to ongoing response, the pembrolizumab was continued until the patient developed pneumonitis after six cycles (42 weeks). Serial imaging showed sustained response until approximately 1 year later, at which time the size of the right adrenal/hepatic mass began to increase. The patient was referred for stereotactic body radiation therapy and his PSA declined once again.

Figure 2.

Radiographic response to bipolar androgen therapy. CT images before bipolar androgen therapy (left) and after (right) show marked interval decreases in the sizes of cervical lymphadenopathy (A) and a retroperitoneal mass (B).

Outcome and follow-up

The patient continued on BAT and off immunotherapy for 4 more months. While on BAT, he experienced an increase in his quality of life, with improved energy, reduction in pain and leg swelling due to metastatic adenopathy, and reduction in his hormonal side effects. His pneumonitis resolved completely with a steroid taper. He continued to play golf and perform independent activities of daily living. He remained in remission for 13 months after starting BAT. Unfortunately, the patient later experienced worsening abdominal pain and was found to have progressive disease in multiple soft tissue sites including rapidly progressive mesenteric and peritoneal masses, pancreatic metastases and bulky abdominal lymphadenopathy. BAT was discontinued and enzalutamide was initiated to determine if BAT may have resensitised his tumour to AR inhibition as per the TRANSFORMER Study.³ The disease continued to progress, with a PSA rise from 0.6 ng/mL to 16 ng/mL and increasing bulky and symptomatic soft tissue metastases. The patient transitioned to hospice care. He passed away 21 months after starting BAT and 23 years after his initial diagnosis.

Discussion

We present the case of a patient with a germline CHEK2-associated mCRPC who progressed through several lines of therapy as follows: brachytherapy, external beam radiotherapy, ADT with leuprolide, combined androgen blockade with bicalutamide, sipuleucel-T, surgical resection, experimental histone methyltransferase inhibition (lirametostat) with enzalutamide, olaparib, cabazitaxel and carboplatin with palliative radiation therapy. After two decades of relapse and progression, he responded to BAT with pembrolizumab consolidation. Patients with mCRPC refractory to therapy are often considered to have survival times less than 6 months. This patient’s progression free survival of 13 months and overall survival of 21 months from the time of BAT/pembrolizumab eighth-line therapy initiation are remarkable and motivated this case report. This is significantly out of proportion to what was anticipated prior to this therapy, where the patient had exhausted treatment options and had an expected survival of less than 6 months and was offered hospice care as an alternative.

We speculate that his germline CHEK2 mutation may have contributed to this therapeutic success with BAT. CHEK2 has been implicated in early-onset and familial prostate cancer.¹³ This pathogenicity is reflected in our patient, who was diagnosed with aggressive disease in his 40s and had two brothers diagnosed with prostate cancer in their 50s. CHEK2 encodes a checkpoint kinase (CHK2) that mediates cellular responses to double-strand DNA breaks.¹⁴ BAT has been shown to induce similar DNA damage in prostate cancer cells.^{15 16} This mechanism may account for the success of BAT in patients with DNA repair-deficient mCRPC, which was recently described in a retrospective analysis of three BAT trials.¹⁷ Of the 15 patients identified in this analysis who had DNA sequencing data and a ‘deep response’ to BAT, all 15 had pathogenic mutations in TP53 or homologous recombination DNA repair genes. CHEK2 mutations were not described in this cohort, but the CHK2 protein is an important sensor in these DNA damage response pathways.¹⁸ Our patient’s preceding radiation therapy may have acted synergistically with BAT in causing double-strand DNA breaks, further exploiting this CHEK2-mediated DNA repair deficiency. Our patient had a low tumour mutation burden and lacked microsatellite instability, and he would not have been predicted to have responded to pembrolizumab monotherapy. The notion of combining BAT with anti-PD-1 immunotherapy was inspired and modelled after the ongoing COMBAT trial of BAT plus nivolumab (NCT03554317), and only prospective controlled trials can in the future establish whether PD-1 inhibition plus BAT can provide more durable disease control over BAT alone.

Another mechanism by which the CHEK2 mutation may have enhanced the efficacy of BAT is via interactions with the AR. A novel regulatory relationship between CHEK2 and the AR has also been described, in which CHEK2 knockdown leads to increased transcription of androgen-related genes.⁹ In the context of this negative feedback loop, inactivating CHEK2 mutations could increase AR signalling and dependence and sensitise prostate cancer cells to BAT. Given that the mechanism of action of BAT is incompletely understood, this remains speculative, and further mechanistic studies are warranted.

After responding to BAT, our patient was given consolidation therapy with pembrolizumab. This decision was made based on results of the COMBAT trial,¹² in which men with mCRPC had favourable responses to BAT and concurrent nivolumab therapy. An earlier case series had investigated ‘extreme responses’ to immunotherapy after BAT and found that two of three such patients carried mutations in homologous recombination DNA repair genes.¹⁹ The mechanism of this association requires further investigation but may be related to increased PD-L1 expression in cancer cells with DNA double-strand breaks²⁰ or pathways by which DNA damage induces immune responses, such as activation of the STING pathway by cytosolic DNA fragments.²¹ Single case report outcomes should not be used as evidence of efficacy for treatment regimes. This case supports further research into long-term and large-scale outcome data.

Patient’s perspective

It was more than 20 years ago when I went for a routine annual physical that my employer required. As I was leaving, [a family member] told me to get my PSA checked. I had no idea what that entailed but I made the request to the doctor. He responded that for a man in his 40’s, a PSA test was not required, but he would perform the test if I desired. A couple of days after the physical the doctor called and told me my PSA was elevated. He thought this was likely a lab error, and that I needed to have it retested. The retest confirmed an elevated PSA of 10.4 ng/ml. I had a biopsy, and a few days later the urologist called and informed me I had prostate cancer. The urologist met with [me and a family member] to explain the staging and the various treatment options: surgery, radiation, hormone therapy, cryotherapy, and watchful waiting. I had no symptoms or warnings of any prostate problems–the diagnosis came as a complete shock.

After much soul searching and research, we selected brachytherapy and external beam radiation. I tolerated the treatment very well and was able to work fulltime throughout. The treatment appeared successful as my PSA continually dropped during six-month routine follow-up visits. That is until approximately 7 ½ years after treatment when the PSA started to rise. Imaging and follow-up biopsies were inconclusive, but the PSA continued to rise. Since that time, I have undergone numerous standard-of-care treatments–in fact, all standard of care treatment for which I was eligible. These treatments included: ADT, bicalutamide, sipuleucel-T, ADT/abiraterone, ADT/enzalutamide, olaparib, docetaxel, cabazitaxel, cabazitaxel plus carboplatin, and radiation and surgery of metastases in lymph nodes. I was found to have a hereditary CHEK2 mutation which guided the use of olaparib; however, this treatment did not work for me.

In addition to standard of care treatments, I participated in multiple clinical trials including a research surgery after sipuleucel-T and the combination of an EZH2 inhibitor plus enzalutamide. Despite efforts to control the disease, I had my left ureter, left kidney, and a small portion of my bladder surgically removed due to metastatic involvement. Standard of care treatments and eligible trials continued.

[A family member] and I met with my oncologist after the last chemo treatment had failed. The situation was bleak! My oncologist explained there were no more standard of care options available for me, and I did not qualify for any further ongoing trials at [the current cancer center]. When given this information, [a family member] asked what that meant. My doctor provided my one and only ever prognosis-I had never asked. I was told that I had 2-6 months. That was difficult to hear. After 20+ years of fighting this disease, however, it was not unexpected. My PSA had risen to an all-time high of 94 ng/mL and my cancer was growing rapidly. I was running out of options.

The one encouraging thing that came out of the meeting was my doctor suggested I look at other world class cancer treatment centers for other possible trials. [A family member] started looking into palliative care options while [another family member] and I investigated treatments. We looked at mushroom therapy and various holistic treatment facilities domestic and abroad. It may sound foolish, but desperate times called for desperate measures. We ultimately decided to contact [the other cancer center] and arranged a meeting with a research oncologist.

The meeting was very interesting and, it turns out, lifesaving. The doctor introduced me to Bipolar Androgen Therapy (BAT), in which a high dose of testosterone is administered every 28 days while continuing [leuprolide] injections every six months. He explained this treatment as paradoxical to traditional thinking and treatment of advanced prostate cancer. He indicated some 200 patients had been treated with this therapy at [the cancer center] and approximately 30% had moderately favorable response. He gave me hope and something to think about.

I did not qualify for a trial but BAT was still offered. I set up a meeting with my local oncologist and spoke to him about the BAT treatment. He had heard of it but was not a strong proponent. He indicated it could potentially accelerate the disease and thus had significant risks and was not yet recommended by any national or international guidelines as a treatment. From my perspective I felt I had little to lose.

I decided to give BAT a try. I set up an appointment for the first injection. Other than the [long drive to get the treatment] it was painless-a blood draw and a shot in the butt. The results were virtually immediate: when I went for the second shot my PSA had dropped by half. In 3 months it had dropped to 1.5 ng/ml, an amazing result and a further sign of hope.

After this dramatic improvement, I had a CT scan. The tumors had shrunk considerably. Based on the results, we continued the treatment protocol with ongoing ADT plus testosterone injections. The BAT treatment continued on a four-week schedule. A CT and bone scan were performed, there was no evidence of disease: the tumors were gone!

Although tumors were later seen in my adrenal glands, I continued BAT and had 5 treatments of stereotactic body radiotherapy (SBRT). My PSA dropped over the next several months to 0.69 ng/ml. The most current PSA obtained was 0.81 ng/ml.

Throughout my journey, I have been blessed with tremendous support. This includes a medical team of highly competent and outstanding professionals, and physical and mental support by my wonderful [family member], whom I lost to cancer. Without their strength and guidance I know I would not be here today. The same could be said about [other family members], and too many friends to name here that have supported me every step of the journey.

Currently I feel great, I can spend quality time with my family and my friends, I continue to work, and I’m able to hit the links a couple of times per week. All journeys come to an end, and mine would have been long ago if not for the support I have received from family, friends, and medical professionals every step of the way.

Learning points

• The backbone of prostate cancer therapy is androgen deprivation and blockade, though several other strategies have been developed in recent years.

• Bipolar androgen therapy (BAT) is a novel paradoxical approach whereby patients are treated with cycles of androgen deprivation and supraphysiological doses of exogenous testosterone.

• BAT is thought to induce DNA damage in cancer cells and sensitise them to anti-androgen therapies.

• An important subset of patients with prostate cancer have predisposing genetic mutations, one of which is in the DNA repair mediator CHEK2.

• Patients with DNA repair deficiencies and metastatic castration-resistant prostate cancer may have enhanced sensitivity to BAT, though further research is needed before recommending this therapy.

Ethics statements

Patient consent for publication

Obtained.