Olanzapine for Chemotherapy related Anorexia in GI and Lung Cancer patients
Olanzapine is an atypical antipsychotic in use since 1991 for treatment primarily in schizophrenia and bipolar disorder. It has a higher affinity for 5HT2a serotonin receptors than D2 Dopamine receptors. Weight gain was one of the main adverse effects as an antipsychotic. In 2016, Navari et al. established the role of Olanzapine 10mg orally daily along with standard antiemetics for highly emetogenic chemotherapy with nearly 30% improvement in acute and late chemotherapy induced nausea and vomiting. Subsequent studies have proven effectiveness of lesser dose (5mg) and also in pediatric cancer patients. In advanced cancers, anorexia is a significant symptom in 30% -80% patients often aggravated by chemotherapy. Different drugs like appetite stimulants and practices like exercise, yoga etc., is being used with varying results.
Role of Olanzapine in improving anorexia in cancer patients has been studied recently by Indian researchers from Jawaharlal Institute of Postgraduate Medicine, Education and Research (JIPMER), Pondicherry. This study was published by Lakshmi Sandhya et al. in Journal of Clinical oncology (Randomized Double-Blind Placebo-Controlled Study of Olanzapine for Chemotherapy-Related Anorexia in Patients with Locally Advanced or Metastatic Gastric, Hepatopancreaticobiliary, and Lung Cancer. (doi: 10.1200/jco. 22.01997). This was conducted as a prospective randomised double blinded placebo-controlled study in adults over 18yrs with locally advanced or metastatic gastric, hepatopancreaticobiliary (HPB) and lung cancers who were treatment naïve.
Olanzapine 2.5mg once a day for 12 weeks or placebo was given along with chemotherapy with dietary advice and standard nutritional assessment provided to both groups. Proportion of patients with weight gain more than 5% and improvement in appetite as assessed by various tools (VAS/FAAC ACS) were primary objective and secondary objective included quality of life, chemotherapy toxicity etc.
Total of 124 patients were randomised into olanzapine (63) and placebo (61) arms, majority had metastatic gastric cancer and lung cancer. Weight gain of 5% or more was found significantly better in the olanzapine arm (60%) when compared to placebo (9%). Appetite also was better in olanzapine arm by both VAS (43% vs 13%) and FAACT ACS (22% cs 4%) assessments. Better nutritional status, QOL and lesser chemotoxicity was seen in olanzapine arm with no significant side effects.
This study concludes that daily olanzapine at low dose is inexpensive and well tolerated intervention that significantly improves appetite and weight gain in newly diagnosed patients started on chemotherapy. This is a practice changing trial for cancer patients globally as cancer associated cachexia and anorexia is of universal concern. Especially in India where baseline severe malnutrition can be seen in more than 50% of cancer patients as seen in various studies. Often this situation worsens when they get initiated on chemotherapy. The cost of Olanzapine 2.5mg per day for 12 weeks is Rs 500 which makes it a much cheaper and effective alternative than the other medications being prescribed for this clinical condition.
Can Artificial intelligence help in identifying individuals who are prone to oral cancers?
Oral cavity cancer is the most common cancer in males in India. More than 1,35,000 new oral cancer cases are detected every year. Sankarnarayana et al. in his randomised controlled study in four Indian district has demonstrated significant impact of screening in high risk individuals in prevention of oral cavity cancers. Though India has achieved WHO recommended doctor to population ratio (1:1000) in 2018 but oncologist to population ratio still needs improvement. Joshi et al. demonstrated that on an average every cancer patient has to travel for more than 10 kms and there is a considerable delay of 7 months before they reach the concerned oncologist. Therefore, with the dearth of oncologist and infrastructure, if we can identify high risk individuals who are susceptible to oral cancers then it will help triage huge number of high risk population to be screened. The study led by John Adeoye titled Explainable ensemble learning model improves identification of candidates for oral cancer screening was recently published in Oral Oncology journal (https://doi.org/10.1016/j.oraloncology. 2022.106278). This study trained and constructed meta-classifiers with comprehensive risk factors information for eleven supervised learning algorithm predicting probability. These probabilities were further weighted and aggregated by the meta-classifiers. A web-based questionnaire consisting of comprehensive risk factors was administered by calibrated interviewers for screening each patient. Data consisted of of twenty-seven risk factors including tobacco use, alcohol consumption, betel nut chewing, second hand smoking, dietary history, frequency of dental visits, family history and medical history. Also, expired carbon monoxide levels were determined in all patients as surrogate marker of smoking or passive smoking.
This retrospective study had high internal validation recall (0.83), specificity (0.86) and AUROC (0.85) all upwards of 80% for detecting oral cancer. It also built algorithm to identify individuals who were at high risk of oral potentially malignant disorder (OPMD). It had high validation recall (0.92) and AUROC curve (0.76) but moderate specificity (<75%) for OPMD. This means AI with the above algorithm has better recall validation as compared to crude method of just identifying patients who consumed tobacco vis-à-vis who didn’t. Similarly, a retrospective study led by Warin et al. based on oral photographic images used DenseNet121 and faster R-CNN algorithm to diagnose oral potentially malignant lesion with acceptable results.
Artificial intelligence (AI) is the use of algorithms or decision making based on the collective experience of the database accumulated over the period of time. Developing algorithm to identify individuals with high susceptibility to oral cancer can go long way to impact the population at large. Cumulative database and deep learning from screening, diagnosis, imaging and histopathological algorithm would significantly improve the rates of early detection with improved accessibility, consequently improving survival. Importantly, this could be done with as simple device as our mobile phones.
