In this paper we consider a stochastic multi-item inventory problem. A retailer sells multiple products with stochastic demand and is replenished periodically from a supplier with ample stock. At each order instant it is decided which product to order and how much to order. For the delivery of the products trucks with a finite capacity are available. The dispatched trucks arrive at the retailer after a constant leadtime and with each truck fixed shipping costs are charged independent on the number of units shipped. Additionally, linear holding and backorder costs at the end of a review period are considered. Since fixed transportation costs are high coordination of orders and full truckload shipments can benefit from economies of scale. We propose a dynamic order-up-to policy where initial ordersizes can be reduced as well as enlarged to create full truckloads. We show how to compute the policy parameters and in a detailed numerical study we compare our policy with a lower bound and an uncoordinated periodic replenishment policy. An excellent cost performance of the proposed policy can be observed when average time between two shipments is not too large and fixed shipping costs are high.