Purpose/Objective(s) Physician surveys of fractionation schedules for palliative radiotherapy (pRT) in Africa have reported low utilization of single fraction or hypofractionated regimens despite studies demonstrating clinical efficacy in a number of different clinical scenarios. However, there is a dearth of data describing the real-world patterns of pRT use in low- and middle-income settings. This study aimed to investigate historical patterns of pRT use in Botswana to determine the rate of single and hypofractionated treatment used in this setting. Materials/Methods Treatment summaries from patients treated with palliative intent between 2015-2022 were retrospectively collected from the treatment planning system at the sole radiation therapy department in Gaborone, Botswana. Data was analyzed using descriptive statistics and chi-squared tests. Results Of 8197 recorded prescribed radiotherapy courses, 1735 were delivered with palliative intent and included for analysis. An average of 217 pRT patients (SD 60; range 127-284) were treated per year. Median patient age was 55 years (IQR 44-66). The most frequently treated histologies were cervix (n = 355), breast (n = 330), prostate (n = 122), esophagus (n = 84), and lung (n = 70). The most frequently treated sites were pelvis (n = 546), spine (n = 322), head and neck (n = 209), brain (n = 136), extremities (n = 121), upper gastrointestinal (n = 80), and chest (n = 68). The most frequently prescribed schedules included 1 fraction (n = 569, 32.8%), 5 fractions (n = 357, 20.6%), 10 fractions (n = 346, 19.9%), and 15 fractions (n = 89, 5.1%). The proportion of these fraction prescriptions was not significantly different by year. The most frequently prescribed doses included 800 cGy (n = 387, 22.3%), 2000 cGy (n = 352, 20.2%), 3000 cGy (n = 338, 19.5%), and 600 cGy (n = 117, 6.74%). Multiple courses were prescribed to 15.2% (n = 263) patients (range 2-10). Among all courses, 82.1% (n = 1424) delivered the dose prescribed. Conclusion In Botswana, over one fifth of prescribed radiotherapy courses were delivered with palliative intent. Primary cancer histologies of cervix, breast, and prostate were most common, reflecting the patterns of cancer incidence in the country. Over half pRT courses were delivered in 5 or fewer fractions and were prescribed to 2000 cGy or less. Most pRT patients received the full prescription dose. Notably, the most commonly prescribed regimen was a single fraction, frequently delivering 800 cGy, suggesting that the rate of use of single fraction pRT has been higher in Botswana compared to rates reported in surveys by other African oncologists.

There was no significant difference in SCCAg among WLWH and women living without HIV. Among patients with elevated SCCAg above normal at baseline, SCCAg was associated with early vs. advanced stage disease. Additionally, there was a significant difference seen in overall survival by two measurement points: baseline SCCAg >7.9 ng/mL and response SCCAg >2.8 ng/mL. SCCAg may be utilized as a biomarker in low-resource settings to refine prognosis. Further studies will be needed to determine utility and validation in predicting recurrence risk and/or lymph node metastases.

Purpose/Objective(s) Timely access to and completion of radiation therapy (RT) are crucial for optimizing clinical outcomes of patients with locally advanced cervical cancer. The cascade of cancer care comprises the steps between initial presentation and treatment initiation. It is unknown which of these steps if any contribute to delay and downstream survival. Multi-disciplinary management has been shown to enhance the likelihood of successful care coordination and delivery across this cascade, but when implemented in limited resource settings it is unknown if multi-disciplinary management initiatives can adequately identify care delivery gaps. This study sought to examine care delivery metrics and outcomes from a newly established multi-disciplinary team clinic (MDT) in a limited resource setting. The analysis benchmarked the timing of steps between diagnosis and RT initiation in patients with locally advanced cervical cancer in Botswana to determine if delay impacted survival. Materials/Methods Between January 2016 and July 2018, 230 patients with biopsy-proven cervical cancer were enrolled in Gaborone, Botswana and followed until November 2019. The number of days between biopsy, pathology, MDT consult, RT consult, simulation, and treatment start were calculated based on dates of care delivery. The association of delay with survival was evaluated using the Kaplan-Meier estimates and multivariable Cox proportional hazards models. Results After diagnosis, 187 (81.3%) patients ultimately received treatment. Times between various steps of the care cascade are detailed in Table 1. Untreated patients had a median of one-month additional delay between biopsy and pathology result (57 vs. 25 days, p=0.003). Univariate analysis showed patients treated with curative intent (EQD2 >65 Gy) had worse overall survival when delay between simulation and RT start was > 12 days (p=0.048). Median survival was 2 vs. 4.6 years for patients with and without delay >12 days, respectively. This association trended toward but did not meet statistical significance on multivariate analysis after controlling for HIV status and stage (HR 2.35, 95% CI 0.95 – 5.85, P=0.07). HIV status was not associated with delay at any point. Conclusion Treatment delay exists in all steps of the care cascade for patients with cervical cancer in Botswana. Delay of >12 days between radiation simulation and initiation was found to negatively impact survival in patients treated with curative intent. This study also demonstrates the feasibility of utilizing the resources of an MDT care model to quantify care delivery benchmarks in a low-resource setting, resulting in identification of actionable targets to improve care delivery and outcomes.

Purpose/Objective(s)Cervical cancer (CC) remains a significant disease burden for women in low- and middle-income countries (LMICs) where HIV infections remain highly prevalent, even in settings with widely available antiretroviral therapy (ART). This study investigates immune profiles of CC patients with and without HIV infection to identify immune correlates of survival and treatment response.Materials/MethodsWomen with CC undergoing chemoradiation (CRT) with curative intent were enrolled in Gaborone Private Hospital, Botswana under the "Ipabablele" protocol (Grant #1U54 CA190158-01). Patient demographics, clinical characteristics, treatment characteristics, and blood samples were collected after informed consent was obtained. Multiparameter flow cytometry (FC) was performed on longitudinal peripheral blood mononuclear cell (PBMC) samples. Frequency of T cell subtypes and cytokine markers were analyzed at distinct points during CRT including prior to initiation (IT), at the completion (EOT), and three months after (M3) CRT. Logistic regression analysis identified immune markers that differed by HIV status and correlated with survival or treatment response (TS), defined as squamous cell carcinoma antigen below 2.2 ng/mL.ResultsThis study analyzed samples from 131 women (HIV+ n=89, HIV- n=42). On laboratory blood tests done at the time of enrollment, the HIV+ patients had a median baseline CD4 count of 454 cells/μL (n=88, IQR=275-590 cells/μL). On FC analysis of PBMCs, HIV+ women had significantly lower CD4 frequency at IT than HIV- women (52.5% vs 72.0%, p<0.001). Between IT and M3, there were no significant changes in CD4 frequency (52.5% to 50.9%) or CD8 frequency (39.9% to 41.4%) in HIV+ women. However, HIV- women demonstrated a significant decrease in CD4 frequency (72% to 60.55%, p<0.001) and increase in CD8 frequency (20.9% to 31.5%, p<0.001) by M3. Both cohorts underwent similar T cell activation during CRT marked by loss of CCR7 and increases in CD57 and IFNγ expression. Significant (p<0.05) findings on logistic regression analyses demonstrated that survival in HIV+ women was associated with the presence of CD8 subsets lacking CD57 and CD28 at IT and EOT. In HIV- women, poor survival was associated with IL-2 and IFNγ expression on CD4 and CD8 cells at IT and positive treatment response correlated with IL-2 expressing CD4 T cells and proinflammatory CD8 T cells at M3.ConclusionThis study demonstrates that CRT induces peripheral T cell differentiation and activation as well as a distinctive cytokine profile in all CC patients. However, a subset of these changes in the immune repertoire differs by HIV status. Patients with HIV also do not experience significant reduction in CD4 frequency while on treatment as compared to uninfected women. These results suggest that in women with well-managed disease, HIV infection may differentially impact immune response to CRT and in turn may reflect broader clinical outcomes including treatment response and survival.

This survey of international respondents indicated considerable interest in the RPA in LMIC settings. Implementation must be tailored to variations in perceived benefits and barriers may vary by provider role, practice location, and infrastructural resources.

Cervical cancer is the leading cause of cancer death among women in Sub-Saharan Africa. Although chemoradiation (CRT) is the standard of care for locally invasive cervical cancer, data on its benefit over radiation (RT) alone in patients with stage IIIB disease has been inconsistent. A recent randomized controlled study focusing on IIIB cervical cancers showed significant survival benefit of CRT over RT in this population. The benefit of CRT over RT has not yet been evaluated in HIV-infected patients, who have a four-fold increased incidence of cervical cancer. We prospectively evaluated treatment outcomes among patients with stage IIIB cervical cancer to assess the survival benefit of CRT vs. RT alone based on HIV status in a resource-limited setting with a high incidence of HIV. Our hypothesis was that there is a significant survival benefit of CRT over RT in treating HIV-infected patients with stage IIIB cervical cancer. Patients with stage IIIB cervical cancer treated with CRT or RT in Botswana were enrolled in a prospective observational cohort study from 2013 to 2018. Survival was evaluated using the Kaplan-Meier method, and associations with survival were analyzed using Cox proportional hazards modeling. Of 187 study patients who were enrolled, 118 (63%) were HIV-positive, with a median CD4 count of 445 (IQR: 319-657). Ninety-six patients (51.3%) received a total EQD2 ≥70 Gy, and roughly half (48%) were prescribed concurrent cisplatin chemotherapy (most commonly 35-40 mg/m2 over 2-4 cycles). When divided by treatment type, patients treated with CRT had improved 2-year OS when compared with those receiving RT alone (59% vs. 41%, p = .005). HIV status was not associated with worse survival outcomes (p = .527). Among HIV-infected patients, those receiving CRT also had significantly higher 2-year OS when compared to those receiving RT alone (60% CRT vs. 38%, p = .022). Other factors associated with improved survival were: total EQD2 ≥70 Gy (HR 0.53, p = .001), receipt of brachytherapy (HR 0.42, p<.001), and complete response at end-of-treatment (HR 0.45, p<.001). Importantly, total RT dose modified the effect of chemotherapy on OS. For patients who received a total RT dose <70 Gy (n = 91), adjusted OS rates were significantly higher among those treated with CRT (adjusted HR 0.47; 95% CI 0.26 - 0.87). In contrast, for patients who received a total EQD2 ≥70 Gy (n = 96), chemotherapy was not significantly associated with OS (adjusted HR 0.97; 95% CI 0.52-1.80). Optimal management of all women with stage IIIB cervical cancer, including those who are HIV-infected, should entail CRT. Addition of chemotherapy directly contributes to improved OS in HIV-infected and uninfected patients. In our study, as expected, the effect of CRT over OS is significantly modified by the RT dose.

Challenges of providing cancer care in Botswana include longer times to diagnosis and limited treatment resources. BOTSOGO (Botswana Oncology Global Outreach) was established in 2011 as a collaboration between Massachusetts General Hospital (MGH) in Boston, MA, USA and Botswana Harvard AIDS Institute in Gaborone, Botswana. Our mission is to improve cancer care in Botswana at the levels of education, research and direct clinical improvements, by means of an ongoing productive, bilateral partnership. A forum was needed for specialists on both sides to review cases, learn about real-time cancer care issues faced in Botswana, and discuss best practices for diagnosis and treatment. The Tumor Board model, in which a multidisciplinary team meets regularly to review cases, is widely accepted as improving cancer outcomes. Since February 2012, BOTSOGO has held a monthly Tumor Board in which MGH and two oncology centers in Botswana (Gaborone and Francistown) connect via web conferencing. A recent cancer case is presented by the treating team in Botswana, and disease experts from MGH discuss pathology, diagnosis and recommended treatment strategies. We promote open and respectful discussion, allowing clinicians from both sides to speak candidly and to highlight any opportunities for improvement. As cancer care is a team effort, the audience comprises attending physicians, trainees, nurses, social workers, and other caregivers. As of February 2020, we have held 79 Tumor Boards, with average attendance of 75 in Gaborone, 50 in Francistown and 15 at MGH. The most frequent cancers presented include cervical, breast and prostate cancers, sarcomas, and lymphomas, with special sessions on Ethical Issues in Oncology; Religious Factors Affecting Patient Decisions; and the Role of Nurse Advocacy in Patient Care. The August 2016 session, “A 41 year old female with sexual dysfunction following treatment for cervical cancer,” led to the implementation in 2017 of a program where women receiving radiation for vaginal cancers are provided dilators and education to treat vaginal stenosis, a frequent side effect. Since 2016, we offer Continuing Medical Education (CME) credits through Harvard Medical School for attendees, with 904 CME credits awarded as of August 2019. Evaluations received to date in the current 2019-2020 academic year (07/2019 – 1/2020; n = 305) report: 83.3% thought the presentation and discussion level was ‘just right;’ 86.6% found the session was ‘very likely’ or ‘somewhat likely’ to change their work; 97.1% of attendees would ‘definitely’ or ‘probably’ recommend attending the BOTSOGO tumor board. Tumor Board discussions are complex and free-ranging. We promote a collaborative environment, making time for questions, comments and insights from the participants in Gaborone and Francistown. The Tumor Board continues to anchor the BOTSOGO partnership.

Head and neck cancer (HNC) is the fifth most common cancer in sub-Saharan Africa (SSA), a region with hyperendemic HIV infection. Individuals with HIV have higher rates of HNC, however the effect of HIV infection on treatment toxicity and oncologic outcomes is not well-characterized. The objective of this study was to prospectively evaluate patterns of oncologic care and outcomes for HNC patients with or without HIV infection in Botswana, a resource-limited SSA country that has one of the highest rates of HIV infection in the world. Patients with HNC attending an oncology clinic were enrolled in a prospective observational cohort registry in Gaborone, Botswana from 2015 through 2019. Clinical characteristics were analyzed to identify factors associated with oncologic treatment and outcomes. Overall survival (OS) was evaluated via Kaplan-Meier and associations with survival and toxicity were analyzed via Cox proportional hazards and logistic regression analyses, respectively. Secondary analysis by HIV infection status was then performed. A total of 149 patients were enrolled with a median follow-up of 23 months. The most common HNC anatomical subsites were oral cavity (37%), larynx (9%), salivary gland (9%), oropharynx (9%), nasopharynx (9%), and orbit (9%). Patients were likely to present with advanced disease (60% of patients had T4 primary disease) and have a long interval from pathologic diagnosis to the start of radiation treatment (RT; median time of 2.5 months [IQR: 1.6-5.4 months]). Only 29% of patients received definitive RT, 19% received chemotherapy, and 8% received surgery as part of their management plan. Median overall survival (OS) was 36.2 months (95% CI: 20.6 – 51.8 months) with a 2-year OS of 58%. Low hemoglobin (<12 g/dL) was associated with poorer survival on multivariable analysis (HR 2.74, p = 0.001). Grade ≥ 3 toxicity rate with RT was 30% and associated with mucosal sub-site (OR 4.04, p = 0.03) and low BMI (<20 kg/m 2) [OR 6.04, p = .012]. Of the 149 patients, 59 were HIV-infected (40%). Most of the HIV-infected patients were on antiretroviral therapy (85%) and had suppressed viral loads (90% ≤400 copies/mL) with a median CD4 count of 400 cells/mm 3 (IQR: 237-584). HIV status was not associated with increased time to initiation of definitive RT (p = 0.440), decreased receipt of definitive RT (p = 0.137), worse survival (p = 0.606), or increased toxicity (p = 0.527). Despite access to government care, delayed presentation remains a significant issue for HNC patients in Botswana. A disproportionate number of patients with HNC are infected with HIV, but HIV status does not adversely impact survival or toxicity outcomes for these patients and should not preclude definitive management.

HIV infection is associated with an increased risk of vulvar cancer development. However, limited data is available on presentation and treatment of patients with both vulvar cancer and HIV, particularly in regions with high prevalence of HIV such as Sub-Saharan Africa. In the current analysis, we aim to describe the presentation, treatment modalities, and clinical outcomes of women with vulvar cancer in Botswana. We prospectively enrolled vulvar cancer patients treated with surgery, radiation (RT), and/or chemoradiotherapy (CRT) between Jan 2015-July 2018 at Princess Marina Hospital in Botswana. Demographics, clinical characteristics, and survival of patients treated with surgery, chemo, and/or RT were recorded. We enrolled 82 women. The median age was 43.0 years (IQR=38.0-47.0). Seventy-three patients (89%) were living with HIV. The median baseline CD4 count was 379.5 cells/uL (IQR=232.5-617.5) with median time of 9.5 years (IQR=5.0-12.0) on anti-retroviral treatment (ART). Twenty-eight (34.1%) women presented with early stage 0/I/II disease and 42 (51.2%) with late stage III/IV disease. Twenty patients had surgery as part of their treatment and of these patients, 11 received surgery alone. Patients who had surgery alone more frequently had early stage cancer with 91% (n=10) being treated for stage 0/I/II disease. Of the 56 patients who received RT, 27 had CRT, 20 had RT alone, 6 had RT and surgery, and 6 had CRT and surgery. Late stage disease was more common in patients treated with CRT (70%, n=14) or RT alone (59%, n=16). Complete response was observed in 20 women (27.0%), partial response in 27 women (36.5%), and no response/stable disease in 3 women (4.0%). At the end of data collection, 70 women (85.4%) were alive and 12 women (14.6%) were deceased. Median survival time for all patients, calculated as time between date of pathology reported and date of death or most recent follow up, was 16.8 months (IQR=11.4-25.8). For those that received surgery alone (n=11, 13.4%), median survival was 22.8 months (IQR=17.0-25.3). The 20 patients who received CRT had a median survival time of 16.9 months (IQR=11.9-24.3). Over half (n=11, 55%) of patients who were treated with CRT received 2 cycles of chemotherapy. Twenty-seven patients (32.9%) received RT alone with medial survival of 13.7 months (IQR=8.3-28.2). The average RT dose for those who received RT alone was 5500 cGy (IQR=3700-5900) compared to 5940 cGy (IQR=5600-6100) for those in the CRT group. Women in Botswana who presented with vulvar cancer had a high prevalence of HIV infection that was well-controlled with long-term ART. As per international guidelines, women presenting with early stages (stage 0/I/II) were more likely to receive surgical treatment while patients with advance stages (stage III/IV) were more frequently treated with CRT or RT alone. HIV status did not play a significant role in the decision of treatment modality.

Quality breast cancer (BCa) treatment requires coordinated multidisciplinary care, which can be challenging in resource constrained settings. We sought to describe predictors of receipt of indicated radiation therapy (RT) in Botswana, a middle-income country with publicly funded oncology services and high HIV prevalence. Using the Botswana Prospective Cancer Cohort (BPCC), we identified women with stage I-III BCa treated with mastectomy (MTX) or breast-conserving surgery (BCS) between 2010 – 2015 who had indications for RT (BCS, pN+, pT3–4, or MTX with positive margins). Women who received incomplete resection or neoadjuvant therapy were excluded. RT receipt was defined as completion of ≥ 45 Gy to the breast/chest wall ± regional nodes (LN). To identify predictors of RT receipt, we used univariate and multivariable logistic regression. Covariates included HIV status, education level, income, marital status, distance from RT facility and number of RT indications. We identified 345 women with BCa captured in the BPCC. Of 154 women with pathologic stage I-III, 129 had ≥ 1 RT indications (stage I, 0.8%; II, 29%; III, 49%). Median age was 50.2 (interquartile range [IQR] 42.6 – 60.2) and 26% were HIV-infected, 61% HIV-uninfected, and 13% had unknown HIV status. Most had MTX (87%) and 13% had BCS. Overall, 83% had adjuvant chemotherapy starting at a median 7.1 months from surgery (IQR 4.5 – 10.5). Of 129 women with RT indications, 83 (64%) received RT. Median time to RT from surgery was 8.3 months (IQR 7.0 – 9.5); doses ranged from 45 – 50 Gy in 2 – 2.5 Gy per fraction with or without 10-Gy boost (67% had boost); 90% had nodal RT. Median number of pathologically examined LNs was 7 (IQR 4 – 11) for women receiving RT vs. 5 (IQR 1 – 11) for those who did not; 12% of women who received RT had 0 LNs examined vs. 24% who did not (P = 0.08). Type of surgery, adjuvant chemotherapy receipt or timing, number of RT indications, pT- or pN-stage, margins, age, HIV status were not significantly associated with RT use on unadjusted analyses. RT receipt was associated with senior secondary (grade 12) or higher education (37% vs. 13%, P = 0.004) and trend for married status (41% vs. 24%, P = 0.052), income > $100 monthly (P = 0.051) and HER2 overexpression (25% vs. 9%, P = 0.063). In adjusted analyses, having senior secondary or higher education (odds ratio [OR] 3.23, 95% CI 1.02 to 10), P = 0.045) and known HIV status (HIV-uninfected vs. HIV unknown, OR 3.85, P = 0.024; HIV-uninfected vs. HIV-infected OR 0.76, P = 0.589) was associated with increased odds of receiving RT. Women who lived within 1.5 hours from the RT center were more likely to receive RT (OR 2.01, 95% CI 0.88 – 4.59), but this was not significant (P= 0.096). Among women with localized breast cancer in Botswana, one-third of those with indications for RT do not receive it. Interventions targeting women with low education, those needing to travel far distances, and less engaged in routine care are needed to address RT underutilization and improve outcomes.

Cervical cancer is the leading cause of cancer death for women in Sub-Saharan Africa and chemoradiation (CRT) is the accepted curative treatment globally. HIV-infected vs. uninfected women with cervical cancer who are able to initiate radical CRT have been shown to have similar outcomes. However, it is unknown what factors are associated with the ability to initiate curative CRT in women with cervical cancer in a low resource setting such as Botswana. We prospectively enrolled cervical cancer patients between July 2013-January 2015 who anticipated to received radical CRT at the only radiation oncology facility in Botswana. Demographic and clinical characteristics were compared between those who initiated radical CRT vs. those who didn't initiate radical CRT (i.e received radiation alone or palliative radiation) using univariate (UVA) and multivariate (MVA) logistic regression. 182 women were enrolled during the study period and 143 women (79%) initiated radical CRT. One hundred and twenty patients (66%) were HIV-infected. On UVA, women who initiated radical CRT were more likely to be older (age 40-59 years 55% [n=79/143] vs. 36% [n=14/39],p=0.02), married/partnered (27%[n=38/143] vs. 13% [5/39], p=0.08), have lower stage disease (IIIA/B 29%[41/143] vs. 62%[n=24/39], p>0.001), have lower baseline creatinine umol/l (median, interquartile range[IQR], 51[42-59] vs. 67[48-95], p=0.01), higher baseline hemoglobin (Hb) g/dl (median IQR 11.2[9.3-12.2] vs. 8.9 [7.8-10.7],p=0.004). There was no significant difference detected in distance from facility, delays in seeking care, performance status or HIV status. On MVA, adjusting for age, marital status, HIV and stage, Hb at start of treatment (Odds ratio[OR] 1.38, 95% CI 1.08-1.72, p=0.009) and creatinine and at start of treatment (OR 0.96, 95% CI 0.94-0.98,p=0.004) were significantly associated with initiating radical CRT. On sub-analysis, using MVA linear regression, HIV (creatinine, p=0.06, Hb p=0.03) and stage (creatinine p=0.08, Hb p>0.001) were strongest predictors for higher creatinine and lower Hb at baseline. Initiation of radical CRT in Botswana is associated with baseline Hb and creatinine at time of presentation for treatment and not HIV status. HIV-infection and stage are surrogates for Hb and creatinine at baseline. This underlines significance of development of supportive care for better access to blood transfusions for advanced stage patients and greater focus on underlying myelosuppression and renal dysfunction particularly in HIV infected patients.