Real-World Outcomes in Patients with Advanced Penile Squamous Cell Carcinoma Receiving Immune Checkpoint Inhibitors: A Single Institution Experience

Penile squamous cell carcinoma (pSCC) is a rare and aggressive malignancy in the advanced setting. In the United States, the incidence is less than 1 per 100,000 and represents approximately 2100 cases annually and less than 1% of cancers.^[1,2] Approximately half of cases relapse or are refractory to standard of care platinum-based chemotherapy.^[3,4] Five-year survival rates for localized, regional, and distant metastases are 79%, 51%, and 9%, respectively, highlighting a need for effective therapeutic modalities for advanced disease.

Numerous risk factors involving chronic inflammation and lifestyle factors are associated with developing pSCC. These include smoking, human papillomavirus infection (HPV), nodal invasion or metastases, socioeconomic factors, phimosis, and poor penile hygiene.^[5] Notably, HPV infection is implicated in 30–50% of pSCC cases as well as in head and neck, cervical, anal, vulvar, and vaginal cancers.^[5] Although HPV/p16-associated head and neck cancers have been associated with improved outcomes with immune checkpoint inhibitors (ICIs),^[6] there has been little success achieved with immunotherapy with an unselected approach in advanced pSCC. However, several studies indicate immunotherapy may confer survival benefits in selected patients with HPV-positive pSCC.^[7–9] There is no consensus regarding the prognostic role of this marker nor is it consistently evaluated in clinical practice.^[2,5]

Treatment of advanced pSCC is multimodal and composed primarily of surgical resection for curative intent and either neoadjuvant or adjuvant platinum-based chemotherapy or chemoradiation.^[10–12] ICIs such as pembrolizumab are indicated as a second-line therapy for pSCC with elevated tumor mutational burden (TMB) ≥ 10, mismatch repair–deficient (dMMR), microsatellite instability-high, or on progression following treatment if no alternative agents are available. In a selected approach, ICIs have been proposed as a viable therapeutic modality.^[13–15] We present a single-center retrospective study with a focus on real-world outcomes, examination of immune-related adverse events (irAEs), and impact of inflammatory biomarkers on treatment response in patients treated with ICIs for advanced pSCC. We also highlight the clinical and biomarker characteristics of three patients with ICI response.

Patient records with a diagnosis of pSCC were queried based on International Classification of Diseases (ICD-9 and ICD-10) codes at the Winship Cancer Institute of Emory University from 2012–2023. Institutional review board approval was obtained, and informed consent was not required due to the retrospective nature of the study. Data collected included demographics, disease histology and stage, treatments, outcomes, and labs that included mainly a baseline complete blood count (CBC).

Patients with a diagnosis of advanced pSCC and received at least one cycle of ICIs were included in the study. Diagnosis was based on histological confirmation. The study included patients who were treated as advanced pSCC based on a constellation of clinical, imaging, and histopathology findings. Notably, 14 patients with biopsy-proven squamous cell penile carcinoma were submitted from our institution to the Global Society of Rare Genitourinary Tumors report of which 10 were eligible for inclusion.^[16] We also included two patients who were treated based on pSCC guidelines in which biopsy showed poorly differentiated SCC versus carcinoma of pelvic origin with unknown primary origin (e.g., penile vs perineal). Normal lab ranges included absolute neutrophil count (0.91–5.53 × 10³ cells/µL), absolute lymphocyte count (0.72–3.29 × 10³ cells/µL), absolute monocyte count (0.14–0.71 × 10³ cells/µL) and absolute eosinophil count (≤0.36 × 10³ cells/µL), lactate dehydrogenase (LDH) (140–271 units/L), and C-reactive protein (CRP) (<10 mg/L). HPV and programmed death-ligand 1 (PD-L1) status were determined by testing or immunohistochemistry. Immune-related adverse events (irAEs) were graded based on Common Terminology Criteria for Adverse Events criteria and were captured through the clinical notes.^[17] TMB measurements were performed based on FoundationCDx and CARIS testing.

Statistical analysis was conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC). The significance level was set at p < 0.05. Demographic and clinical characteristics were summarized using descriptive statistics. Clinical outcomes included overall survival (OS), which was calculated from time of ICI initiation to death or last follow-up. Progression-free survival (PFS) was calculated from ICI initiation to clinical progression, radiographic progression based on Response Evaluation Criteria in Solid Tumors (RECIST) 1.1, or death, whichever event occurred first. Clinical benefit was defined as achieving complete response (CR), partial response (PR), or stable disease (SD). Patient deaths were confirmed by reviewing the electronic health record as well as the Georgia online obituary database. The Kaplan-Meier method was used to generate PFS and OS curves, and survival curves were compared between groups using the log-rank test. The univariate analysis (UVA) between covariates and survival outcomes was assessed using Cox proportional hazards model. Hazard ratios with 95% CIs were reported. The median follow-up was calculated using the reverse Kaplan-Meier method. Multivariable analysis (MVA) was not performed because of limited sample size. We calculated baseline neutrophil-to-lymphocyte (NLR), monocyte-to-lymphocyte (MLR), platelet-to-lymphocyte (PLR), and neutrophil-to-eosinophil ratios (NER) based on the CBC before ICI initiation. These measurements were subsequently dichotomized as high versus low subgroups based on optimal cutoffs determined by first examining the nonlinear relationship between numerical data and PFS/OS and then categorizing it as high/low. Optimal cutoffs were set at high/low defined as NLR ≥ 6.74 (42.8%), MLR ≥ 1.13 (14.3%), PLR ≥ 346.5 (33.3%), and NER ≥ 49.4 (47.6%). An additional OS curve was created for NLR with cutoffs of > 5 versus NLR ≤ 5.

Overall, the study had a total of 21 patients (Table 1). Approximately 71% of patients were White and 29% were Black. Median age at ICI initiation was 56 years (38–76 years). Most patients had Eastern Cooperative Oncology Group (ECOG) performance status ≥ 2 (61.9%). Nine patients had a smoking history (42.9%). At diagnosis, 4.8%, 33.3%, and 61.9% were stage 2, 3, and 4 respectively; eight patients had initial distal metastases. Most common histological subtype was keratinizing (42.9%). Of the patients who were treated initially with surgical resection, median tumor size was 6.75 cm (0.3–19.5 cm) and surgical margins were negative in 41.2% of patients. Ten patients had poorly differentiated disease on histology (47.6%). Nine patients had lymphovascular invasion, and nine patients had perineural invasion. Eight patients received total penectomy (38.1%) and 16 patients (76.2%) received external beam radiation therapy (EBRT) during their disease course.

Nearly all patients received prior chemotherapy (90.5%). Nine had received neoadjuvant therapy (42.9%) (Table 1). Median follow-up was 8.4 months and median time to ICI treatment from diagnosis was 15.6 months. Median time to ICI initiation from diagnosis was 15.6 months. ICI was first line (1L) in 1 patient (4.8%), 2L in 13 (61.9%), 3L in 6 (28.6%), and 4L in 1 (4.8%). Monotherapy with nivolumab or pembrolizumab was the most common regimen; others included combination ICI, cemipilimab, or immunotherapy within clinical trial. Most patients discontinued ICI due to disease progression. Three patients (14.3%) developed grade 3/4 ICI-related adverse events related to skin, gastrointestinal, renal, and cytopenias. Two patients received steroids as treatment.

Median OS and PFS for the overall cohort were 8.2 months and 1.9 months respectively (Figs. 1 and 2). There were no differences with respect to race or other demographic variables (Fig. 3). Five patients (23.8%, 3 PR, 2 SD) achieved clinical benefit and 76.2% progressed after ICI. Only 14.3% were alive at recent follow-up. Most common sites of disease progression included local invasion (66.7%), lymphadenopathy (57.1%), retroperitoneum (47.6%), lung (42.9%), and bone (42.9%) (Table 1).

For biomarkers before ICI initiation, median level of CRP was 43.1 µg/mL (0–201.9 µg/mL), LDH was 140.5 units/L (99–414 units/L), NLR was 6.87 (2.49–45.46), MLR was 0.71, PLR was 273.5, and NER was 49.4 (Table 1). Of patients with available data, 7 of 11 patients (63.6%) were PD-L1+ and 8 of 9 (89%) patients were HPV+.

Elevated NLR, MLR, PLR, and NER were associated with shorter OS. Patients were categorized into high or low subgroups based on optimal cutoff as follows: high defined as NLR ≥ 6.74 (42.8%), MLR ≥ 1.13 (14.3%), PLR ≥ 346.5 (33.3%), and NER ≥ 49.4 (47.6%).

UVA for OS demonstrated decreased survival in patients with elevated NLR with HR 10.0 (95% CI, 2.73–52.2, p = 0.002) (Figs. 4a, b), MLR with HR 4.83 (95% CI, 1.13–17.3; p = 0.026), PLR with HR 3.28 (95% CI, 1.11–9.95; p = 0.036), NER with HR 3.04 (95% CI, 1.07–9.59; p = 0.049) (Table 2). Acknowledging a limited sample size and data, HPV/p16+ (Fig. 5), PD-L1 positivity (Fig. 6), LDH, and CRP were not associated with survival.

Elevated NLR was associated with shorter PFS (Fig. 7). UVA at optimal cutoff demonstrated HR 4.82 (95% CI, 1.73–13.7; p = 0.003) (Table 2). Acknowledging a limited sample size and data, HPV/p16+, PD-L1 positivity, LDH, and CRP were not associated with progression.

Three patients in this study responded to ICI. The first patient was diagnosed at age 64 when he noticed an abnormality at his urethral meatus and developed obstructive voiding. An ultrasound and CT revealed bilateral enlarged lymph nodes with the largest measuring 2.5 × 1.9 cm. A urologist performed a biopsy of the glans revealing pSCC and referred for partial penectomy with right inguinal dissection. Initial disease stage on diagnosis was stage 4 (T3N3M0). Pathology revealed HPV+, basaloid squamous cell histology, poor differentiation, positive surgical margins, positive lymphovascular invasion, and perineural invasion. He initially received four cycles of TIP (paclitaxel, ifosfamide, and cisplatin) achieving CR; however, he developed progressive disease with bilateral external iliac lymphadenopathy and underwent a nodal dissection and received EBRT and capecitabine with good response. Comprehensive genomic profiling revealed TMB-high (19 mut/mb), MSI-stable, mismatch repair (MMR) proficient, and PD-L1 50%. He subsequently received pembrolizumab for 2 years, achieving PR at the time of analysis. His imaging continues to demonstrate no evidence of disease while off therapy.

The second patient was diagnosed at age 75 when he was noted to have a growing right groin mass while being evaluated for a hip transplant. He received imaging 3 months later revealing multiple pelvic and inguinal nodes. Core needle biopsy revealed invasive squamous cell carcinoma with PD-L1 >50%. Initial disease stage on diagnosis was stage 4 (T3N2M1). Positron emission tomography (PET) scan again revealed the pelvic adenopathy but unremarkable for definitive disease. On discussion with the patient, he underwent TIP chemotherapy and radiation therapy, achieving a PR but developed disease progression 10 months later with retroperitoneal adenopathy and a right upper lobe (RUL) metastatic nodule. He subsequently received pembrolizumab with one cycle with radiographic improvement, achieving PR and resolution of his metastatic RUL involvement. He subsequently deferred further therapies as he developed decreased function and mobility and transitioned to hospice. The patient did not have a longstanding response but had clinical stability with observation alone and continued to attend clinic visits over 16 months and subsequently had disease progression and was admitted to the hospital with significant abdominal pain and swelling in the setting of acute renal failure. Imaging demonstrated significant retroperitoneal adenopathy concerning for recurrent disease, and the patient was transferred to the intensive care unit for renal failure requiring hemodialysis. Comfort care was prioritized after goals of care discussion and the patient died.

The third patient was diagnosed at age 72. He initially developed a slow urinary stream and was treated for benign prostate hyperplasia with minimal improvement. He then underwent a cystoscopy with dilation and biopsy revealing for invasive pSCC. Initial disease stage on diagnosis was stage 3 (T3N1M0). A partial penectomy confirmed this diagnosis with positive margins. The patient was then referred to our urology clinic and a complete penectomy revealed keratinizing histology, HPV+, PD-L1 > 50%, poor differentiation, and negative surgical margins. A postsurgical magnetic resonance imaging scan revealed locoregional occurrence with left pelvic and inguinal lymphadenopathy, osseous involvement, and a right upper lung nodule. Biopsy of the RUL nodule confirmed metastatic disease. Comprehensive genomic testing was unrevealing with low TMB (9 mut/mb), MSI-stable, MMR proficient. He was referred to medical oncology for palliative chemotherapy with four cycles of TIP, achieving PR. He later developed progressive disease demonstrated on PET scan and transitioned to pembrolizumab. The patient has not had evidence of disease recurrence since his last follow-up.

Advanced pSCC is a rare and aggressive malignancy when progressed beyond platinum-based chemotherapy. There has been limited success of immunotherapy with an unselected approach. We report a retrospective analysis reporting an OS of 8.2 months, PFS of 1.9 months, and clinical benefit rate of 23.8%, consistent with other studies. Higher NLR, MLR, PLR, and NER above optimal cutoff were associated with shorter survival. Acknowledging a limited sample size, most clinicopathologic variables including race, HPV, and PD-L1 status were not associated with outcomes. Notably, we highlight three patients with either PD-L1+, dMMR, or MSI-H status who developed therapeutic response with immunotherapy in this real-world study. Our study provides a broad, granular perspective of outcomes in advanced pSCC across a demographically diverse population.

A multicenter, international retrospective study of 92 patients with advanced pSCC demonstrated median OS of 9.8 months (95% CI, 7.7–12.8) and PFS of 3.2 months (95% CI, 2.5–4.2 months) with an overall response rate (ORR) of 13%. Shortened OS and PFS were associated with presence of visceral metastases, ECOG PS ≥ 1, and higher NLR on multivariable analysis.^[16] Patients with lymph node–only involvement demonstrated longer OS than those with visceral and/or bone metastatic disease. Patients with lung metastases had a lower ORR than those without. There were also no significant survival differences comparing HPV+ versus HPV− and PD-L1+ versus PD-L1− patients. Our results were consistent with this analysis. We provide granular patient details and correlate clinical and molecular attributes of patients with ICI response, adding onto the existing literature. HERCULES (LACOG 0218, NCT04224740) is a phase 2 single-arm trial evaluating 37 patients receiving pembrolizumab and platinum-based chemotherapy in treatment-naïve advanced pSCC with confirmed ORR as the primary endpoint.^[18] One patient achieved CR and 12 achieved PR with 39.4% demonstrating confirmed ORR (95% CI, 22.9–57.9). Notably, a confirmed ORR was seen in 66.7% of patients with PD-L1-negative versus 33.5% with PD-L1-positive disease, 75% with high TMB versus 36.4% in low, and 55.6% with HPV16-positive versus 35.0% with negative disease. Twelve of 29 patients were MMR proficient and demonstrated ORR 41.4%. Most frequent genomic aberrations were TP52, CDKN2A, and TERT. Owing to its heterogeneity across primary vs metastatic sites, PD-L1 expression may not be a consistent predictive marker of response. Similarly, a phase 1 study of nivolumab, ipilimumab, and cabozantinib and a phase 2 multicenter basket trial of nivolumab and ipilimumab reported similar response rates.^[19,20] A case series noted one patient with MSI-H demonstrated a durable response to pembrolizumab, whereas two patients with microsatellite stable (MSS) disease had developed radiographic disease progression.^[13]

As seen in head and neck and cervical cancers,^[21,22] HPV infection may prime the tumor immune microenvironment to respond to ICIs, although this mechanism is not clear. A comprehensive tumor sequencing study of 108 patients with advanced pSCC revealed distinct HPV-dependent oncogenic and ICI-based signatures.^[7] Regarding ICI-based markers, 51% were PD-L1+, 10.7% had TMB-H, and 1.1% had MMR deficiency/MSI-H. Our study reported similar PD-L1 proportions.^[23] PD-L1+ proportions were similar across HPV+ versus HPV− populations. TMB-H (30.8%) was found solely in the HPV+ group. Independent genomic signatures were also associated with molecularly distinct tumor profiles with KMT2C and FGF3 mutations exclusively found in HPV+ tumors, whereas CDKN2A mutations were found exclusively in HPV− tumors. A retrospective study consisting of 87 patients compared outcomes in HPV+ versus HPV− patients with advanced pSCC receiving first-line ICI. HPV+ patients demonstrated better OS, PFS, and ORR compared with HPV− patients.^[24] In addition, therapeutic HPV vaccines are being developed and may potentially augment responses to ICI therapy. A phase 2 clinical trial explored efficacy in ICI and HPV-16 vaccine demonstrating a median OS of 17.5 months and ORR of 33% compared with 16% to 22% across ICI alone, suggesting potential synergy.^[25] Although not standardized, HPV status may be beneficial as a potential biomarker for stratifying patients within ICI-based clinical trials.

Rarity of disease implicates significant challenges in trial accrual due to patient enrollment. Large cooperative retrospective studies create opportunities to study rare cancers.^[7] Umbrella trials with molecularly selected patients and large data sets are potential strategies to mitigate this. There are no current FDA–approved molecularly targeted treatments. However, tumor agnostic approaches within precision oncology may shed light on therapeutic advances. Recent studies have revealed a potential association of HPV positivity with Nectin-4 and Trop-2, expanding our current understanding of HPV-driven pathogenesis and a potential role for novel therapies like targeted antibody-drug conjugates.^[26–28] Elevated inflammatory markers are known to be associated with worsened survival outcomes in many malignancies.^[29,30] Several retrospective studies have demonstrated patients with high NLR had significantly decreased cancer-specific survival consistent with our study.^[30,31] Studies investigating the immune microenvironment revealed that higher densities of CD8+ cells and tumor-associated (M2) macrophages may be associated with shorter median OS; however, there were no differences in cell densities across HPV+ and HPV− pSCC.^[32] A study using multiplex immunofluorescence demonstrated changes in macrophage and T lymphocyte density in advanced nodal stages, suggesting immune tolerance and exhaustion as potential mechanisms of ICI resistance. Several tumor sequencing studies report an association of HPV/p16+ with elevated TMB^[7,8,33,34]; however, there is no consensus regarding the prognostic role of this marker nor is it consistently evaluated.

Emerging trials evaluating combinations of immunochemotherapy and/or molecular inhibitors are currently under way. A phase 2 open-label, multicenter study is evaluating efficacy of dostarlimab and niraparib in patients with chemo-refractory advanced pSCC.^[35] This combination is postulated to increase tumor neoantigen release and expression of immune checkpoint–based biomarkers, potentially sensitizing the tumor immune microenvironment to ICI.^[36] A phase 2 multicenter, nonrandomized clinical trial evaluating cemiplimab alone or in combination with cisplatin-based chemotherapy is currently in progress.^[37] PERICLES is a phase 2 single-center trial evaluating efficacy of atezolizumab with or without radiation in 32 patients with penile pSCC; however, did not meet its primary PFS endpoint.^[38] Other notable phase 2 trials investigating immunotherapy outcomes in advanced/chemo-refractory pSCC include NCT03391479 and NCT02721732 as well as basket trials (NCT02834013).

Our work has several limitations. Owing to the study’s retrospective nature, confounding and selection bias may have affected potential associations. The small sample size limited an evaluation of clinicopathologic variables. As several patients were referred to our center for treatment, there was no centralized pathology review. Strengths of our study include the single-center nature, patients being from a similar geographic region of the United States, and ethnic makeover of the study that mirrors metro Atlanta. Moreover, we provide granular clinical and biomarker data describing a favorable response in a select group of patients. These findings may provide guidance in selecting patients with advanced pSCC who are more likely to be ICI-responsive.

Advanced pSCC is a rare and aggressive malignancy with limited treatment options after progression from standard of care platinum-based chemotherapy. We report overall limited efficacy outcomes in patients treated with immunotherapy with an unselected approach consistent with other studies. ICIs may be effective and tolerable in a subset of patients. Elevated NLR may be a potential biomarker for identifying patients who may respond to immunotherapy. Prospective, multicenter collaborative efforts enriched with clinicogenomic and ICI-based biomarkers are required to further validate our findings and may be useful to stratify patients in ICI- and novel agent–based clinical trials.

The authors thank the Global Society of Rare Genitourinary Tumors (GSRGT) for advancing outcomes research in penile cancer. The authors submitted 14 patients, and 10 were included in a separate global analysis by GSRGT.^[16] Portions of the work from this manuscript were presented at Society of Immunotherapy Conference 2022 and ASCO GU Symposium 2023.